How to get through the Ballard Locks: the procedure, the strategy, how the locks work, commercial priority, and the days to avoid

Going through the Ballard Locks: the boater's procedure

Transiting is free, first-come (with commercial priority), open 24/7, and requires no reservation or radio call. The whole system is run by lights and loudspeakers.

Before you go

• Two 50-ft lines, bow and stern, each with a spliced eye — a federal requirement (33 CFR 207.750(b)(5)(v)), and the attendants will use them.
• Fenders on both sides — you will not know which wall you get until you're assigned, and undersized fenders are the most common first-timer mistake.
• Life jackets on during transit is the Corps' strong recommendation: lines, props and crowded chambers are the risk zone.
• Check the live forecast before you leave the dock — when you cast off matters more than most boaters assume: the hour-by-hour medians are on the wait-times page, and what to do about them on the strategy page.

The sequence

• Red light: hold clear or tie to the waiting piers on the south wall (both approaches). Not the north wall.
• Green light: proceed at no more than 2.5 knots and listen — the attendants assign you a chamber, a side and a position by loudspeaker.
• Small lock (the usual recreational route): tie bow and stern to the numbered buttons on the floating guide wall and ride it up or down. No line-tending. Watch that the float doesn't jam; call out if it does. Order is strictly first-come, first-served by arrival — boaters track and respect who showed up when, regardless of where each boat sits in the waiting area. Your place in line is set the moment you arrive; jockeying for position gains nothing and costs goodwill.
• Large lock: no floating wall, and loading order differs — the largest recreational boats are called in first to take the wall and tend lines as the water moves 6–26 feet (a two-person job). Smaller boats are then rafted alongside the big ones, which is often the easy ride: you tie to a hull instead of handling wall lines, and their crew has already done the work. If you're small and get sent to the large lock, it's usually to your advantage.
• Exiting upbound (into the lakes): release the lake-side line last. When the gates open, current runs seaward and a loose bow swings. Downbound, the stern line goes last for the same reason.

Timing it

Measured hour by hour on the wait-times page: noon runs about double the evening, and every figure there is transit time — arrival to clearing the chamber, the lockage included — not queue time alone. Which hour to pick, the commercial windows to dodge, how to read the chamber cycle and why being rigged before the green light matters are the four levers on the strategy page; today's conditions, the tour-boat timetable and approaching tugs are on the live forecast.

After the locks: the bridges, and when they won't open

Sailboats bound for Lake Union or Lake Washington may need bridge openings (Ballard, Fremont, University, Montlake — roughly 30, 44, 44 and 46 ft closed clearance; the Fremont at 30 ft opens most often). Signal ~100 yards out; an opening takes 4–6 minutes. But by federal regulation (33 CFR 117.1051) the draws do not open for vessels under 1,000 tons during weekday road rush hours:

None of these closures apply on weekends or federal holidays — the windows are a weekday phenomenon — with one exception: Columbus Day keeps the Ballard, Fremont and University windows. The front-page calendar draws them over the day when its Seattle bridges button (Ballard, Fremont, University) or its Montlake button is pressed. Two more rules worth knowing: any draw may hold you up to 10 minutes to clear accumulated road traffic, and between 11 pm and 7 am the draws open only with one hour's advance notice: the Ballard/Fremont/University spans through the Fremont drawtender (phone or radio), the Montlake through the bridge at 206-386-4251 or VHF channel 13, an hour ahead; the Montlake Cut's own number is 206-720-3048. Practical upshot — sailboats only (powerboats under the ~30-ft Fremont clearance are unaffected): outbound is where the windows bite hardest, since the bridges stand between your moorage and the locks and gate your departure entirely — mast traffic leaves in waves just after 9 am and 6 pm. Inbound they're a lesser but real factor: you can always lock through, but clear upbound during a window and the Ballard Bridge five minutes upstream will hold you in the canal.

One more bridge, and it is before the locks: the BNSF Salmon Bay railroad bridge sits at mile 0.1 of the Ship Canal, just west of the chambers, and every boat coming in from the Sound passes under it. It is a bascule that stands open for marine traffic and comes down for each train — 30 to 40 a day — and lowered it clears 43 ft at mean high water (about 50 ft at low water), so most cruising sailboats wait while it is down, roughly 12 minutes a train. Only the passenger trains keep a timetable: Sounder and Amtrak (Cascades and the Empire Builder), about six rail miles and twelve minutes from King Street, which puts 14 scheduled closures on a weekday and 6 on a weekend or Sounder holiday. The front-page calendar draws those as thin hatched bands over the day so you can plan a lockage around them. Amtrak trains are followed live (positions from Amtraker), so an Amtrak band and the railroad-bridge line on the forecast card update as the train closes on the span and again once it has crossed, while Sounder publishes no live positions and stays on the timetable. Freight closures — 20 or more a day — are unscheduled and are not shown; treat the drawn bands as the floor. The tender answers on signal (33 CFR 117.1051) and the span is normally up when nothing is on the line.

The Amtrak bands learn from the bridge. The twelve-minute run from King Street is an estimate, and the tracker banks the minute each Amtrak train actually crosses. Once a train number has five banked crossings, its band on every day of the calendar moves to the measured crossing — the median of the minutes it was actually seen on the bridge, whichever timetable minute either side had estimated — and widens to the measured spread (the middle half of those crossings, plus the 12-minute closure). Hover a band and the title says how many crossings it rests on and how far the train typically runs from its timetable minute. Fewer than five and the band stays where the timetable puts it; Sounder is never moved, since it publishes no positions to bank. Freight is still not in it.

Common questions

What do I need to go through the Ballard Locks?

Two 50-foot mooring lines with an eye splice (required by 33 CFR 207.750), fenders rigged on both sides, and life jackets on deck. There is no fee, no reservation, and no VHF call required for recreational boats — the lights and loudspeakers direct you.

Which side do I wait on at the Ballard Locks?

Tie to the waiting piers on the SOUTH wall on either approach if the light is red. Do not tie to the north wall.

How fast can I enter the Ballard Locks?

Guides state a 2.5-knot entry limit; the governing regulation (33 CFR 207.750) caps speed at 4 knots between the guide piers.

Do first-timers get help at the locks?

Yes. Tell the attendants it's your first time and they will talk you through every step. Watching a few lockages from shore first — or on the locks webcams — is the standard advice.

What VHF channel do the Ballard Locks monitor?

The lockmaster monitors VHF channel 13 and answers (206) 783-7000, staffed 24 hours. Recreational boaters normally just follow the lights and loudspeaker instructions.

The fastest way through the Ballard Locks

📊 This is the written guide: the reasoning, the rules and the mechanisms behind them. The live charts for your direction and boat type are in the app — open LockPush and pick the Optimal strategy tab. Planning a day ahead? Tomorrow, hour by hour draws the expected transit for each half hour of the next day.

A note on the numbers: every headline figure on this page is transit time — arrival to clearing the chamber, the lockage itself included. It is not the time you spend waiting: on the wait-times page the median transit is 27 minutes, of which the queue — arrival to doors closing, the part actually spent waiting — is 8 and the chamber cycle 18.

First, understand what you're actually dodging

The single most useful fact in a month of lockage records is this: recreational traffic during daylight is nearly constant, and commercial traffic is not. Measured on the wait-times page: recreational arrivals vary by only about 14% from hour to hour and run flat across the week, while commercial arrivals swing 8.5-fold between the busiest hour and the quietest. In summer, there is no lunchtime crowd and no weekend surge: the recreational fleet is a steady background hum whenever the sun is up. (Transit times do move a little with the day of the week — Tuesday has the lowest estimate — and those figures are in the weekend section below.)

Commercial traffic is the lumpy ingredient — and it explains a paradox in the data: the median transit roughly doubles from the evening to noon (the hour-by-hour figures are under decision 1 below) while recreational arrivals barely move at all. The slow hours are not a recreational rush. They are two things layered onto the steady hum:

• The commercial priority overlay — commercial arrivals cluster 10 am–3 pm rather than spreading evenly, the tour boat runs its two fixed daily lockages, and each one seizes a full small-lock cycle that recreational boats sit out. Measured: a commercial lockage in the 15 minutes before your arrival takes the median transit from 26 minutes to 42.
• Cycle luck — whether you make the chamber cycle you arrived for or sit out the one behind it. Arrive within ten minutes of a door-closing and the median transit is 50 minutes, against 24 for a boat arriving 35 or more minutes in — 2.08 times, at any hour, crowd or no crowd.

Honesty note: this is one month of high summer, and one month of high summer is one month — the measurement window is stated on the wait-times page, and the hour-of-day pattern is untested outside August. We don't claim it holds year-round: off-season recreational demand will thin and likely get lumpier (fair-weather Saturdays vs dead Tuesdays) while the tug fleet keeps its weekday rhythm. This page updates as the archive grows.

Two playbooks: does your boat need bridge openings?

Everything below applies to every boat — but if you carry a mast taller than the Fremont Bridge's ~30-foot closed clearance, you're running a second clock: the weekday bridge windows (7–9 am, 4–6 pm) gate when you can reach or leave the locks at all. Powerboats: use decisions 1–4 below and skip the bridge section. Sailboats: read the bridge section first — on a weekday it constrains your options before transit optimization even starts, and the live forecast now flags bridge-window arrival times with a ⛵ marker.

How the time actually divides

Transit time is queue + chamber. The chamber cycle is fixed physics — the one part of the trip no decision shortens, and in a month of records no boat of any kind got under it. So the entire game is the queue: the median queue quoted above is the overall median, not the good outcome, and it runs to an hour-plus when commercial priority and cycle position both go against you. Four decisions control it.

1. Pick the hour — worth 24 minutes

This is the largest of the four levers and the one most boaters leave to chance. Measured hour by hour on the wait-times page, the median arrival-to-clear time runs 39–44 minutes through the 11 am–2 pm peak, 22 in the 3 pm hour and 20–22 from 6 to 9 pm — 44 at noon against 20 at 8 pm, a 24-minute spread across daylight. That is more than the whole small-lock cycle, and more than any other single decision here except cycle position. For a boat that transits twice a week it is a couple of hours a month spent tied to a waiting wall, bought or lost before you cast off.

What makes it worth measuring rather than guessing is that the shape of the day is not the shape intuition predicts. The obvious model — transit times follow crowds, so read the day the way you would read a car park — fails on both counts. First, there is no recreational crowd to read: arrivals are so nearly flat that the busiest hour and the quietest hour look much the same from the fuel dock. Second, the transit curve has structure in it that watching the waiting area would never reveal. It does not rise and fall in a smooth hump; it steps. Two o'clock and three o'clock sit an hour apart on the clock and more than an entire chamber cycle apart in median transit, and the transitions are sharper than anything in the arrival data that supposedly causes them. Two identical boats an hour apart — same day, same direction, same weather — routinely have completely different afternoons.

The mechanism is the two ingredients from the top of this page, layered and slightly out of phase. Commercial movements are clustered, and each one takes a small-lock cycle that the recreational queue sits out; a run of them builds a backlog that outlives the vessels themselves, so the slow stretch of the day extends past the last commercial hull and then drains at the chamber's own pace, not instantly. Underneath that sits cycle luck: whether you make the next chamber cycle or sit out the one behind it, worth 50 minutes against 24. The good hours are simply the hours in which you are likely to make the cycle you arrived for instead of the one behind it. Because those two effects peak at different points, the day has edges rather than a curve — which is exactly why "go when it looks quiet" is a poor proxy, and why the answer is not the one most people would guess.

The 3 o'clock window. The sharpest edge in the day, and the one worth planning around: the 2 pm hour still sits inside the midday peak and the 3 pm hour is as quick as the evening. The midday commercial cluster ends once the early-afternoon tour boat clears, its backlog drains at the chamber's own pace, and the evening arrival bump hasn't started. The drop holds in both directions — one small chamber serves both, so commercial traffic either way delays everyone and its ending relieves everyone — and it holds at weekends, because the tour boats run daily: the 3 pm hour runs a median of 22 minutes on both kinds of day. (We tested whether the 3:30 pm bridge closures cause the inbound dip: they don't — arrivals don't drop, and the dip is there at weekends, when no bridge windows exist. The weekend 3 pm hour rests on 9 lockages, so treat that half of the test as indicative.) Aim to arrive 2:45–3:30; by 4 pm the median has climbed back out of the window, and it does not return to evening speed until 6 pm. One sailboat exception: inbound and bound past the Montlake Bridge, a 3 pm lockage puts you at Montlake around 4 pm, inside its 3:30–6:30 pm closure (3:30–7 pm September through April) — take the morning instead.

The evening block. Three o'clock is one hour wide. The evening is four: 6 to 9 pm holds the 3 o'clock speed, sustained — so instead of hitting a single hour you can leave the dock when the boat and crew are actually ready, which for most people is the more useful kind of fast. The cost is light: "after 6 pm" means dusk by late August and full dark by October, and nobody wants to work a lock chamber, or raft to a stranger, in the dark. Early carries the same trade in reverse — the 6 am hour is nearly as quick as 3 pm, but the window narrows fast, hour by hour, and by 11 am you are inside the midday peak.

Two things hold whichever hours you settle on. The pattern is close to identical in both directions: one small chamber serves inbound and outbound, so commercial traffic either way delays everyone, and its absence relieves everyone. And weekdays and weekends differ less than you would expect: the shape of the day is the same on both, and at the same hour of daylight a weekend trip runs about 4 minutes longer at the median — a sixth of the 24-minute hour-of-day swing (the next section has the figures). The tour boats run daily and the recreational fleet is flat across the week; what changes at the weekend is commercial exposure, which is lower (decision 2). Worth saying plainly, too: the fastest hours and the most comfortable hours are not always the same — which is the whole argument for 3 pm over the evening once the light goes. A choice that is fast on the records but awkward in practice is not automatically the right one for your boat and crew.

Songs of the locks

Two songs from the docks, each revised so that every number in it is one the record supports.

The Sound's Locks

The Chittenden Anthem, country and folk, a little over a minute.

Little Boat Schedule

A minute long, for the youngest crew: a nursery rhyme with the same figures.

Checked against 37 days of Corps records, 900 recreational transits: outbound at dawn queues a median 4 minutes and clears in 19; inbound at dusk queues 5 and clears in 22, and the reverse pairing is nearly as good; tour and passenger boats are 18% of commercial lockages and almost all run between 11 am and 3 pm; a boat arriving within fifteen minutes behind a commercial vessel clears in 40 against 26 with none near, with a 7-minute queue; midday runs about twice dusk, and at any hour the unlucky tenth takes roughly three times the lucky tenth. The one line the record softened: dawn is a fine time to come in, but dusk is better.

Does the day of the week change your wait?

A little, and not the way most people expect: at the same hour of daylight, a weekend trip runs about 4 minutes longer at the median than a weekday trip (32 minutes against 28, arrival to clear), commercial traffic cannot be what makes them slower, and the fits allow it at most about a third of the size, and on the records it is a Sunday effect: Sunday runs 8 to 10 minutes over the weekday average at the same hour, while Saturday's estimate is 0 (with an interval of about −3 to +7, so a slow Saturday is not ruled out). That is the answer once you compare like with like. The raw comparison shows a slightly smaller gap (29 minutes against 26 across all hours, 31 against 28 in daylight) because weekend boats arrive at a slightly different mix of hours; hold the hour fixed and the gap is 4 minutes either way you measure it. Reweighting the weekend to the weekday mix of arrival hours gives +4 (90% interval 0 to +11 when whole days are resampled), and a regression with a term for every daylight hour of arrival and for direction gives +4.0 (90% interval 0 to +9). Against the 24 minutes the hour of day moves the median, the day of the week is a small lever.

Recreational lockages whose arrival fell in daylight, 30 July – 4 September 2026 (37 days, the last of them the morning of 4 September only: 27 weekdays, 10 weekend days). Median transit, arrival to clear. The bottom row reweights the weekend to the weekday mix of arrival hours. The noon–3 pm weekend cell, at 30 lockages, is the thinnest.

Where the extra minutes are. Not in the queue. Hold the hour fixed and the median queue — arrival to doors closing — is 9 minutes on both kinds of day (the mean is 13 at weekends against 15 on weekdays). The whole weekend difference is chamber time, the doors-closed-to-clear part that no decision shortens. Across the day, not hour-adjusted, the median chamber time is 20 minutes at weekends against 18 on weekdays (6 am to 9 pm; 21 against 18 on the daylight set), and holding the hour fixed every 3-hour band runs 2 to 5 minutes longer at weekends: from 9 am to 3 pm a weekend lockage spends 23–25 minutes in the chamber against 19–20 on a weekday, with 20–33% of weekend lockages taking 30 minutes or more against 7–12% on weekdays. The penalty as measured sits in the middle of the day (+9 to +10 minutes from 9 am to 3 pm) and is small early, late and in the evening (+2 to +3), but those are point estimates: a morning-versus-afternoon test with the regression did not resolve the split at 90% (−3 minutes, interval −10 to +5), and the noon–3 pm weekend cell is 30 lockages that split Saturday 51 / Sunday 41. The records do not count boats per lockage, so the cause is not measured; fuller, more mixed rafts locking through more slowly is the natural reading, and it is a reading.

By day of the week, at the same hour of daylight and against the Monday-to-Friday average: Monday 0 to +1 minute, Tuesday −4, Wednesday −2 to −3, Thursday +1, Friday +4 to +8, Saturday 0, Sunday +8 to +10. Where two figures are given they are the two methods; each day of the week is 5 or 6 calendar days. Once whole days are resampled, two days' intervals clear the weekday average by both methods: Sunday (90% interval +5 to +18 by one method, +3 to +15 by the other) and Friday (+2 to +18, or +1 to +8); Tuesday, the quickest day of the week, clears it by one method (−8 to −1) and touches zero by the other (−6 to 0). With seven days compared at 90%, one of these would be expected to appear by chance. As raw colour rather than the adjusted comparison, every one of the five Sundays ran above the weekday average on the day's own daylight mean (35 to 57 minutes against 34). Saturday's interval (−3 to +7) straddles zero. On these ten days the weekend penalty is a Sunday penalty; a Saturday looks like a weekday on average, but five Saturdays cannot promise any one of them (the 29 August Saturday ran a 41-minute median).

What commercial traffic explains: not the direction, and at most about a third of the size. Commercial lockages run 16.0 a day on weekdays against 9.4 at weekends (13.5 against 8.1 in daylight; 2.5 against 0.6 in the 6–9 am band), so if commercial exposure drove the difference, weekends would be faster. They are not, and adding the count of commercial lockages in the same hour, or in the two hours before arrival, to the regression moves the weekend term by between −1.3 and +0.4 minutes (to +3.8 or +4.3 on the mean, +2.7 or +3.6 on the median; the +2.7 comes from the same-hour count, which also includes commercial lockages that happened after the boat had cleared, so it is the least clean of the four). The forecast's own finer covariate — a commercial lockage in the 15 minutes before you arrive — was not fitted here; the strongest reason commercial traffic cannot explain the weekend is simply that there is less of it at weekends. Tuesday points the same way, though it is only five days and also the lightest recreational day (22.8 lockages a day): it carries the most commercial traffic of any day, 21.8 a day, and has the lowest transit estimate. Commercial traffic still costs you a full cycle when one is in the quarter hour ahead of you (decision 2); it just does not decide the day of the week.

Caveats: 37 days of high summer (30 July – 4 September 2026), 922 recreational and 526 commercial lockages, of which 822 recreational arrivals fell in daylight; ten weekend days (five Saturdays, five Sundays) against 27 weekdays, so the intervals are wide: the pooled weekend difference's interval touches zero on the median (0 to +11, 0 to +9) and crosses it on the mean (−1 to +10), so the +4 is a best estimate, not an established gap; one of the five Sundays (2 August) is Seafair Sunday, and the Sunday figure does not depend on it (the slowest Sunday was 23 August), but with five days that is worth knowing; no federal holiday falls in the window, so "holiday Mondays behave like weekends" is how the forecast classifies them, not a measurement; and the pattern is untested outside summer. The unadjusted weekday and weekend counts are on the wait-times page.

2. Dodge commercial traffic — worth a full cycle

Commercial vessels lock first by federal rule and take the small lock alone, and the Goodtime III's two daily lockages — ~11:24 toward the lakes and ~13:59 toward the Sound, ±6 minutes — run on a timetable: the priority rule and the schedule are on the commercial-priority page. The measured cost: if a commercial vessel locked through in the 15 minutes before you arrived, the median transit was 42 minutes (n=93) against 26 when it had not (n=688). It holds whatever else the day is doing, because the cost is structural rather than crowd-related: a cycle spent on a commercial vessel is a cycle the recreational queue does not get. Give the tour-boat windows twenty minutes of clearance. The live forecast flags scheduled vessels and watches approaching tugs by AIS; the working tug fleet runs weekdays only, so weekend commercial exposure is roughly half.

3. Read the cycle — worth ~26 minutes

The counterintuitive one. Seeing a lockage depart and hustling in behind it is the worst arrival: boats arriving within 10 minutes of a door-closing recorded a median transit of 50 minutes — you've just missed the boat, literally, and a big departing batch usually marks a rush period, not a drained queue. Arrive 35+ minutes after the last cycle in your direction and the median falls to ~24, and the long tail of hour-plus transits largely disappears. The live page's "fastest possible passage" card computes exactly this from the latest records, including which level the chamber is resting at.

4. Be ready to load — worth not losing everything above

To be clear about what readiness buys: at the small lock it does not move you up — arrival order rules, and boaters respect it. What readiness does is protect the place you already have: when your turn comes and you're not rigged, you don't get held for — you forfeit the slot and watch the doors close. Lines rigged bow and stern on both sides (you don't pick your wall), fenders out, crew briefed — before you're called. Fumbling your turn doesn't cost minutes; it costs your direction's next chamber cycle. The large lock loads differently — biggest recreational boats first to take the wall, smaller boats rafted alongside them, usually the easier ride — and the whole procedure, lights to lines to walls, is on the transit guide.

What this looks like in practice

• The ready single-hander. A 30-footer arrives third in line, rigs lines and fenders while idling, and when the second boat is still flaking dock lines at its own turn, the order holds — the scrambler forfeits its slot and the single-hander, ready when called, slides into the vacated place. Nobody cut the line: the unready boat gave its place away. That is the only way readiness ever changes the order — by protecting yours when someone else loses theirs.
• The wrong-side surprise. A crew hangs all four fenders to port because "we always get the south wall," gets assigned starboard, and spends ninety seconds re-rigging while the attendant moves to the next boat. In a full chamber that's the margin between making this lockage and leading the next queue. Fenders on both sides makes the assignment irrelevant.
• The lucky raft. A 22-foot runabout sent to the large lock behind a 48-foot trawler does exactly nothing difficult: comes alongside, hands two lines to the trawler's crew, fenders kiss, done — no wall lines, no tending through a 20-foot drop. The trawler's crew did the work when they took the wall. Small boat, big lock: the easy draw.
• The buried lines. A sailboat arrives at a green light with dock lines still in the lazarette. By the time they're flaked and led, the loudspeaker has moved on and the gates swing shut. That mistake costs the whole chamber cycle you had already earned — the most expensive two minutes of stowage on the Ship Canal.

Sailboats only: the bridges gate your timing — hardest outbound

This section applies to masted boats; powerboats under ~30 ft of air draft slide beneath the Fremont and can ignore it. The Ship Canal draws do not open for recreational vessels on weekdays 7–9 am and 4–6 pm (Ballard/Fremont/University; Montlake's windows are wider — 33 CFR 117.1051).

Outbound (to the Sound) is where this bites hardest: the bridges stand between your moorage and the locks, so during a closure window you cannot even start — your departure is gated before the locks ever see you. The result is visible bunching: outbound mast traffic leaves in waves just after 9 am and just after 6 pm on weekdays, so arriving at the locks at 9:30 puts you inside a small sailboat convoy. Cast off so you clear the last bridge before the window opens... or simply before 7 am.

Inbound it's a lesser but real factor: you can always lock through — the locks never close for rush hour — but if you clear upbound between about 6:40–8:55 am or 3:40–5:55 pm on a weekday, the Ballard Bridge five minutes upstream will make you hold in the canal, having banked nothing. Time the lockage so your bridge arrival lands outside the window. Between 11 pm and 7 am, any opening needs an hour's advance notice to the Fremont drawtender.

And the railroad bridge, on either passage: the Salmon Bay railroad bridge just seaward of the locks is down about 12 minutes for every train, 43 ft closed. The scheduled passenger closures — 14 a weekday, 6 a weekend — are drawn on the front-page calendar as hatched bands; the morning Sounder run (about 7:00–8:45) and the evening one (about 3:50–6:05) stack four trains each, so a masted boat arriving at the Sound side then can lose a chamber cycle waiting for the span. Freight is not on the calendar, so the bands are the least you will meet, not the most.

The compounding math

The first two rows come from the distribution on the wait-times page — the fast quartile begins at about 20 minutes, and no boat of any kind cleared in under 10 — and the levers on this page are how you get onto the fast side of it. The bottom row starts from the measured 50-minute median for a boat arriving within ten minutes of a door-closing; commercial exposure and a fumbled loading stack on top of that. We have quoted tighter best cases here in the past and the records don't support them: a routine passage in single figures is folklore. Same boat, same locks — the spread between the first and last row is pure strategy, and the hour you choose is the largest single lever inside it. Check the live forecast before you cast off; it runs this playbook against the current state of the locks automatically.

Common questions

What is the fastest way to get through the Ballard Locks?

Arrive when a chamber is likely ready in your direction and nothing commercial is close ahead: not within 15 minutes of a commercial lockage, not right after doors close, and rigged to load before you are called. When you arrive matters on the same scale — about 24 minutes of median transit across daylight between the worst hour and the best; the hour-by-hour medians are on the wait-times page.

Is it faster to arrive right after another boat locks through?

No - that's the worst moment. Boats arriving within 10 minutes of a door-closing recorded a median transit of 50 minutes, arrival to clear. Arrive 35+ minutes after the last cycle and the median drops to ~24 - a ratio of 2.08.

Should I avoid the locks when a tour boat is due?

Yes. Commercial vessels take the small lock alone, so one ahead of you costs a full chamber cycle: a commercial lockage in the 15 minutes before your arrival raises the median transit from 26 minutes (n=688) to 42 (n=93). The tour boat's daily timetable is on the commercial-priority page.

Does being first in line help at the Ballard Locks?

It sets your floor, not a shortcut: first in line at a ready chamber still means the full chamber cycle plus minutes of loading, and the chamber cycle is the one part of the transit no decision shortens. The measured chamber and queue times are on the wait-times page.

Is the strategy different on weekends?

Barely. The hour-of-day shape is the same, recreational volume is flat across the week, and commercial exposure is lighter at weekends because the working tug fleet is a weekday operation — the weekday-weekend comparison is in the weekend section above and the counts on the wait-times page. The tour-boat windows still apply daily. The forecast classes holiday Mondays with weekends; no holiday fell in the measured window.

Are weekends slower at the Ballard Locks?

Slightly: at the same hour of daylight a weekend trip runs about 4 minutes longer at the median than a weekday trip (32 minutes against 28, arrival to clear; a best estimate from ten weekend days whose 90% interval runs from 0 to about +10), the extra is chamber time rather than queue, and commercial traffic — lower at weekends — cannot be what makes them slower, and the fits allow it at most about a third of the size. On the records it is a Sunday effect: Sunday runs 8 to 10 minutes over the weekday average at the same hour by two methods (90% interval roughly +3 to +18), while Saturday's estimate is 0 (interval about −3 to +7, so a slow Saturday is not ruled out); and against the 24 minutes the hour of day moves the median it is a small lever.

How the Ballard Locks actually work

The Hiram M. Chittenden Locks connect saltwater Puget Sound to the freshwater Lake Washington Ship Canal. Designed by Major Hiram M. Chittenden, first vessel through August 3, 1916, formally opened July 4, 1917 — 63 years to the day after the canal was first proposed at an 1854 picnic. On the National Register of Historic Places since 1978, and by vessel count the busiest lock complex in the United States by vessel count. The Corps quotes ~45,000 a year, but its own operational records have not reached that in a decade: 30,195 vessels in 2025, 77% of them recreational, and a peak of 41,767 in 2023. Traffic is down 22% since 2016. Plus well over a million visitors.

Two chambers, three jobs

The Corps' three official purposes: hold the lakes at 20–22 ft above sea level, block saltwater intrusion, and move boats across the difference. In normal operation the small lock does almost all the work — around 91% of lockages, by local reckoning. That share is not something we can measure: the lockage records never say which chamber served a boat. The one chamber fact the records do settle is that a barge tow physically cannot fit the small lock, and tows are about 5% of lockages (n=63).

The lift, and why tide barely matters to your wait

Because the lakes are fixed and the Sound tides, the lift varies from ~6 ft (extreme high tide) to 26 ft (extreme low). Intuition says high tide = shorter cycle. Measurement says otherwise: joining 1,176 lockages to NOAA's Seattle gauge, tide height moves chamber time by ~0.07 minutes per foot — a couple of minutes across the entire tidal range. Filling isn't the bottleneck; gates, valves and loading boats are.

Saltwater defense and salmon

• A hinged saltwater barrier on the large lock floor blocks the dense salt wedge; deep-draft vessels (over 14 ft) signal one long, two short blasts to have it lowered. A drain system returns ~160 cfs of salt water downstream, mostly reused to supply the fish ladder.
• The fish ladder (1916, rebuilt 1976 to 21 weirs) carries salmon past the dam — one of the few ladders anywhere at a salt/fresh boundary, with a public underwater viewing gallery. Sockeye peak in July, Chinook late August, coho late September.
• Six spillway tainter gates (32 × 12 ft each) manage lake level to a hundredth of a foot; in spring, smolt flumes shoot young salmon safely past the dam.

What this means for your transit

Everything above is the mechanism behind the numbers on the wait-times page: a fixed mechanical chamber cycle, a queue driven by demand and commercial priority, and a small lock that is the real server for recreational Seattle.

Common questions

How big are the two Ballard Locks chambers?

The large lock is 80 by 825 feet with an intermediate gate so mid-size traffic doesn't cycle the whole chamber; the small lock is 30 by 150 feet and handles most recreational traffic. Source: US Army Corps of Engineers.

How much does the water level change in the Ballard Locks?

The lakes are held at 20–22 feet above sea level while Puget Sound rises and falls with the tide, so the lift is typically around 20 feet — as much as 26 feet at an extreme low tide, as little as ~6 feet at a very high one.

Does the tide change how long a lockage takes?

Measurably, almost not at all. Joining 1,176 lockage records to NOAA's Seattle tide gauge and controlling for time of day: about 0.07 minutes of chamber time per foot of tide. Gates, valves and boat loading dominate the cycle, not the water height.

Why do the Ballard Locks exist?

Three official jobs: hold Lake Washington and Lake Union at 20–22 feet above sea level, keep Puget Sound saltwater out of the fresh lakes, and move boats between the two levels. They opened July 4, 1917.

Are the Ballard Locks really the busiest in the country?

By vessel count, yes — the Corps itself calls them the nation's busiest at about 30,000 vessels a year (2025; ~45,000 is an older figure), most of them recreational. By tonnage they are tiny compared to Mississippi River locks.

Commercial traffic at the Ballard Locks: who goes first, and when

Commercial priority is the single biggest surprise in recreational wait times. It isn't a courtesy — it's federal regulation (33 CFR 207.750): government and scheduled passenger vessels, then commercial merchant vessels, then recreational craft in arrival order. And in the small lock, commercial vessels don't share the chamber with recreational boats at all: one commercial arrival ahead of you costs one full cycle.

The one vessel that runs on a timetable

Across a month of Corps records, exactly one vessel behaves like a scheduled service: the Goodtime III tour boat — toward the lakes at ~11:24 and toward the Sound at ~13:59, daily, with a ±6-minute interquartile spread, on 29 of 31 days. Those twenty-minute windows are the most predictable commercial-priority events of the day, and the live forecast flags them automatically.

Everyone else runs on jobs and tides

The barge tows (Wasp, Island Chief, the Coastal gravel fleet) are the slow movers — they need the large chamber and heavy line work, hence 26–32-minute chamber times. They are also strictly Monday–Friday, so weekend commercial traffic runs at about half the weekday level, leaving mostly the tour boats — the weekday and weekend counts are on the wait-times page.

What LockPush does with this

The forecast detects timetabled vessels automatically from the rolling records (so a winter schedule change tracks itself), watches approaching commercial traffic live by AIS on the Sound side, and adds one full cycle to your expected wait when one is inbound — because that's what the no-sharing rule costs you. The commercial arrival histograms on the Data analysis tab show the full hour-by-hour pattern.

How a vessel near your arrival is priced. The forecast cards give every commercial vessel an occupancy window: the window opens 10 minutes before the vessel arrives, because the lock holds for commercial traffic once it has radioed, and closes when it clears — its chamber time (measured for the hull where the records allow, else about 20 minutes light and 30 with a barge), plus about 6 minutes for the empty chamber to come back when it is bound your way. A vessel whose window covers your arrival is in your way; one whose window has closed, or has not yet opened, is not. A commercial vessel bound your way costs you what is left of its lockage plus the chamber coming back to your level: its chamber time less however long it has been in, plus the 6 minutes for the empty chamber to return — measured: with a commercial vessel already in the chamber bound your way, boats queued a median 13 minutes beyond what was left of its lockage, 7 of which they would have queued with the locks clear. So one 25 minutes ahead of you with a 30-minute lockage still counts, and is priced 5 minutes for it to clear plus 6 for the chamber. One bound against you can cost nothing, because it leaves the chamber at your level and you go straight in as it comes out — unless another commercial vessel is bound your way, which then sets the price. And since the lock will often hold for a vessel that has radioed from ten minutes out, one due up to 10 minutes behind you bound your way is priced as if you were held for it, and marked likely rather than certain. A price under 8 minutes leaves the card clear but still shows the minutes. The figures are measured where the archive has the case — a vessel already in the chamber either way, one due behind you the other way — and estimated from that arithmetic where it does not.

Common questions

Who has priority at the Ballard Locks?

By federal regulation (33 CFR 207.750): US government and City of Seattle vessels and scheduled passenger vessels first, then commercial merchant vessels, then recreational craft in order of arrival, with log tows last.

Do commercial and recreational boats share the Ballard Locks?

Not the small lock — commercial vessels take a cycle to themselves there. In practice that means a commercial arrival ahead of you costs one full cycle, roughly 29 minutes.

What time does the Goodtime III go through the locks?

Like a train: toward the lakes about 11:24 and toward the Sound about 13:59, daily, within ±6 minutes — measured across 29 of 31 days of Corps records.

Is there commercial traffic at the locks on weekends?

About half the weekday level — the working tugs (Wasp, Island Chief, the gravel fleet) run Monday–Friday; the daily tour boats run all week. Weekday and weekend counts are on the wait-times page.

Why did that tug take so long in the lock?

Barge tows need the large chamber and careful line work: a measured median chamber time of 29 minutes, well above the all-vessel chamber median on the wait-times page, plus waiting for that specific chamber.

The days to avoid: boat events that jam the Ballard Locks

Recreational traffic through the locks is normally a steady hum — but a handful of calendar days break the pattern, when an event funnels the whole Puget Sound fleet through one 30-foot-wide chamber. These are those days.

The big five, in traffic order

Worth knowing, second tier

• Wooden Boat Festival, Port Townsend (weekend after Labor Day; next Sep 11–13, 2026): classic-boat exodus Thursday–Friday, return Sunday–Monday.
• Race weekends (Race to the Straits early May, Swiftsure late May): sailboat fleets out Friday, back Sunday–Monday.
• Fishing fleet departures (spring, staggered): Ballard's commercial fleet heads for Alaska across several weeks — not one spike day, but extra large-vessel priority lockages through April–May.
• Any hot summer Friday evening / Sunday afternoon: not an event, but the weekly mini-exodus and return follow the same shape in miniature.

Common questions

What is the busiest day of the year at the Ballard Locks?

Opening Day of Boating Season — the first Saturday of May — when the parade sends hundreds of boats through. Experienced boaters simply don't transit that day.

Is Seafair weekend busy at the locks?

Yes, measurably: the busiest days in LockPush's recorded archive (46 lockages/day, Jul 30 - Aug 2, 2026) were Seafair weekend, as Lake Washington-bound traffic funnels through.

Are these event days accounted for in the LockPush forecast?

Yes. Every listed day is marked in the app's charts and classified into the weekend traffic regime (or heavier) by the forecast, alongside federal holidays.