It’s hard to believe that Route 40, now the most obvious way to move between Ballard, Fremont, and South Lake Union, didn’t exist only a few years ago. It’s younger than STB, and yet it’s hard to imagine transit travel in that corner of the city without it. It draws 13,000 riders a day while being late on 20% of trips across the entire day.

Well now, it’s the 40’s turn to get the Move Seattle treatment. The objective is to cut peak travel time by 5 to 10% with priority treatments, which would make the route take about the same amount of time all day. The project is supposed to complete in 2024.

There are no details on specific projects, except that they’re focusing on pretty obvious chokepoints: SLU, Fremont, Ballard, etc. (see below) For now, Metro is asking you to take the survey and/or drop in at one of their open houses:

Tuesday, March 17: 12 – 1:30 PM
Y at the Cascade People’s Center
309 Pontius Ave N

Wednesday, March 18: 5 – 6:30 PM
Leif Erikson Hall
2245 NW 57th St

Thursday, March 19: 5 – 6:30 PM
Loyal Heights Community Center
2101 NW 77th St

Conversion to RapidRide is not part of the project. Read the fact sheet here.

41 Replies to “SDOT plans Route 40 improvements”

  1. One persistent and annoying issue is NB 40s starting their trips late. I will catch the 40 at 3rd/Madison (3rd stop of the route) and they are often 3-4 minutes late already. This is fairly common, and occurs both at rush hour and midday. Midday is most perplexing to me, since the bus traffic on 3rd is low enough to not cause bus-induced delays.

    On a route with 15 minute headways off-peak and 5-6 minute headways at peak, these initial delays can lead to overcrowding and bunching.

    1. the SB 40s starting at Northgate usually start late too. I commute to work (SLU) and I get on at Meridian & 105th – early in the route…..the bus usually arrives a few minutes late.

  2. Best news could be bringing bus drivers in on the project from the get-go. Looking at the route map for the 40, it’s surprising the delays aren’t longer.

    One thing I do remember. Block or two on Ninth southbound a couple of blocks north of Fairview, fewer than a dozen parking spaces blocked a long enough lane on the west side of the street to hold a line of buses for fifteen minutes.

    Considering the amount of territory served by the 40, there could be, what, a hundred places along the map where conversion of a lane from car parking to bus running could save a fortune in operating time, by the clock and the pocketbook?

    Any talk of trolleywire?

    Mark Dublin

    Mark Dublin

    1. does transfer point just mean “another bus line also serves this stop”? if so, these look reasonable to me as a resident of the area and regular 40 rider. I haven’t confirmed every single one but overall they are where I would expect.

    2. It does. The 15, 17, and 29 all overlap before the SB bus gets to the D line, for example. I do not know about the rest

    3. OK, that could be it. It includes shared stops as well as transfers, and includes buses that only run during rush hour. But even in that case, it is flawed. They forgot 105th and Greenwood (for the transfer to the 5), along with the northern transfer point for the 28. Nor is it only same stop transfers. The 40 bus stop on 105th and Aurora is only served by the 40, as is the stop under the Ballard Bridge.

      The best explanation is that they meant both shared stop transfers *and* crossing transfers, but simply forgot a couple places. It still needs to be fixed.

      It should also have a second symbol (e. g. “shared stop”, as opposed to just “transfer”).

    4. RossB: yes, too many and missing the one with Route 5 at Greenwood Avenue North.

  3. It seems like a stop diet is in order. Some of these would be easy picking. Getting rid of the stops at 78th and 83rd for example, would create better, more consistent stop spacing. This would make for a faster, more consistent travel.

    1. eh, as a rider, I think the stops are OK overall. if nobody is getting on or off there, the driver bypasses them. The major slowdowns are in the trafficy area at the south end of Lake Union (labeled Westlake on the map). Southbound approaching Roy St, we will sit through a light cycle trying to get into that bus-only lane because one car is blocking it just slightly, since they turned into the car lane but got stuck in gridlock. I think the number 1 thing to make this route go faster would be enforcing traffic laws at Mercer: give out massive tickets for blocking the intersection for a few days to try to disincentivize people from squeezing through a yellow light when they can’t fit on the other side.

      1. In the past, I agreed with you. But the problem with stops that are used infrequently is that they destroy reliability. The 40 is a long route, where reliability is every bit important as overall speed. You not only have the potential for lots of missed transfers, but bus bunching as well. The way you address both is by delaying the route when no one uses those stops, which puts us back to the beginning, with a slower bus route.

      2. Stop diets are ablist against disabled folks. That’s why I think that when they introduce the new MLK rapidride, they should keep the 7 as a local, maybe run less frequently.

    2. There are still slowdowns associated with drivers checking whether somebody is there, or stopping because it looked like somebody wanted this bus when actually they wanted another route. The stops on 24th may not be prone to this but it occurs on other routes. And the presence of extra stops encourages riders to spread out so the bus ends up stopping more times than if the stops were consolidated.

      1. I don’t disagree that it causes a little slowdown, but compared to the gridlock near Mercer and coming out of downtown northbound during rush hour, I think skipping stops would have minimal impact. Making the left turn to get to the stop in front of the SLU Whole Foods is extremely aggravatingly slow because of gridlock.

      2. Yeah, there is no way I would say that stop spacing is the worst problem with the 40. Not even close. But since Metro takes this “let’s focus on this route” approach, you might as well solve the minor problem, while fixing the major one. This is what they are doing with the 7, and it makes sense to do that with the 40.

  4. I ride the 40 almost every day from downtown Ballard. IMO this route would be helped massively by providing an ‘X’ express version at rush hour.

    There are a ton of boardings on this route at the hubs (downtown Ballard, downtown Fremont, SLU, 3rd Ave), and VERY LITTLE in the in-between zones (Ballard bridge, Frelard, Westlake).

    Having an express version would have a huge impact on most of the 40’s riders and wouldn’t require any infrastructure changes.

    1. Route 40 Express Ballard to Downtown via Fremont and South Lake Union might not have to be limited to peak hour. Understand serious part of Ballard burned down awhile ago. Anything at all dynamic about the rebuild?

      Also, could be Norse legend, but remember some talk about extending the South Lake Union division of our streetcar system to Ballard via Westlake and Fremont. Need not be an “either-or”. Fully reserved signal-preempted lanes could handle joint operations.

      Leading to further blending of transit technology and culture. Would like to see Metro get close to the new truly world-class Nordic Museum for a permanent arrangement like whatever Swedish is for “Sister Cities” relationship with the joint rail and bus transit system in Gothenburg.

      And while we’re on that coast, maybe Oslo, Norway. Both reachable by same fast electric train from the basement of the airport in Copenhagen, which is an easy transfer at Reykjavik from Icelandair out of Sea-Tac. Overnight.

      Which effort I think can easily lead to working ties with a certain program regarding Link. Think about it. Whichever station location is chosen for Ballard, overseas-exchange tours can start with Link ride from there to Sea-Tac.

      Another thing Seattle’s got in common with both cities above: Being forced by the notorious Low Bid dominion to buy new vehicles from same source as our Breda fleet. With same results- except harder on the Norse because they knew better but had to buy them anyhow.

      Would be great if, thanks to Ballard, we could all join forces to fight back against a repeat.

      Mark Dublin

    2. There may not be a ton of boardings near the Ballard Bridge but it is a useful stop at NW Leary & 15th for the fact of the UW Ballard clinic right there. It’s also a transfer point for those on the D Line.

      1. Metro wouldn’t skip that stop. A “40X” would be a great connection for the D-Line and 15X. Maybe just convert the “peak-only” runs to “X” runs. And make sure they’re all 60 footers.

    3. I think there are enough express buses, I just think they need to be managed a little better. The 18 should stop in the heart of Ballard, and run more often (every ten minutes, if not more). If anything, it is backwards — it skips the stops that are likely most crowded (those between 65th and Leary). The 28 is an express — gathering riders in Fremont willing to walk up the hill — and seems OK. The rest of the route (from Fremont to Westlake) seems OK, with relatively big stop spacing. I don’t think you would gain much by running an express along Westlake — not as much as you would gain by fixing the big traffic bottlenecks (which would hurt expresses as much as the regular buses).

      Other than improving the 18 — making it make a few more stops in exchange for far more frequent service — I think the routing is OK.

      1. The 18X is great for commuters in Ballard headed downtown, but does nothing for SLU or Frelard bound riders. Plus, there are folks from Shoreline along Aurora and Bitter Lake who would probably prefer a five stops express to central Ballard to riding the E to 46th (good!) then the 44 west on Market (not terrible, but not wonderful).

        You’re usually an advocate for a many-to-many network. Peak expresses on this route would link three urban centers in a useful arc.

      2. That would mean a bus that runs every 6 minutes (like today), but half the time it makes all the stops, while the other half of the time it makes only half of them. It doesn’t make sense to watch a bus go by and catch the express, so only those that can time their trip actually benefit. Transferring riders only have a 50/50 chance of benefiting. Unless they are going a long distance, the speed difference is minimal anyway (trips from Aurora to Ballard wouldn’t be much faster). Then there are the riders that are skipped. They have 12 minutes frequency (a clear degradation). You also have a much higher chance of bus bunching. Even if they didn’t actually bunch, they would likely be much closer to each other, meaning effective frequency gets worse for a lot of riders.

        It makes way more sense to just do a little stop reduction (on the entire route, for every 40), along with adding stops to the 18 so that it can serve more people, and run more often. The largest block of riders in Ballard are right along 24th from 65th to Leary. At best some of them have the infrequent 17, while others have the infrequent 18 (and some are stuck in between the routes, and only have the 40). These riders would have a much more frequent bus that not only makes a much faster trip to downtown, but also serves Interbay, Expedia, Belltown and similar areas (not served by the 40).

        Ideally you find money in the budget to add the extra service to the 18. But even if you took it from the 40 (so that it maxed out at 7 minutes, instead of 6) it would likely be worth it, as that would be a minor frequency reduction for those trips, but a major frequency improvement for the others.

      3. Huh? The 18 does stop in the heart of Ballard. It stops at 22nd and Market. The problem is that the 17/18 don’t serve riders going to Fremont or SLU, which in my experience is almost everyone who uses the 40.

        The problem during rush-hour is that the 40 is very crowded with people, so even the most sparsely-used stop will be requested by at least one person (it often is literally just one person using the Westlake/Frelard stops). Because the bus is so crowded it then takes at least a minute for the single person to fight their way to the door and get off, and then everyone who got off to give them space to get back on.

        FWIW, I think we should do both. Fix traffic bottlenecks AND create a 40X at rush-hour.

      4. The 18 does stop in the heart of Ballard. It stops at 22nd and Market.

        Sorry, that was a bad summary of the area I was talking about. My point is that the 18 should have more stops between 65th and Market. The 18 is backwards. It makes every stop north of 65th, even though there are fewer people there. It doesn’t stop at 57th or 61st. On the 40 those account for about 1200 riders a day. No wonder, given the tremendous growth there. (Which explains why the 18 is structured that way — years ago there weren’t that many riders on that part of 24th, now they account for the bulk of it).

        The problem during rush-hour is that the 40 is very crowded with people, so even the most sparsely-used stop will be requested by at least one person (it often is literally just one person using the Westlake/Frelard stops). Because the bus is so crowded it then takes at least a minute for the single person to fight their way to the door and get off, and then everyone who got off to give them space to get back on.

        Yes, and what I’m saying is that if you added stops to the 18 and increased its frequency, more people would use it, instead of the crowded 40, in much the same way that people use the 301 instead of the crowded E.

        The problem is that the 17/18 don’t serve riders going to Fremont or SLU, which in my experience is almost everyone who uses the 40.

        On an outbound (northbound) bus, about 2,700 people board downtown, south of 6th and Blanchard. Between there and the Fremont Bridge, it picks up a bit less than 2,000. It is impossible to determine where those downtown riders get off, but just by looking at the total number who get off before Ballard, it is clear that there are lots of people taking the 40 from Ballard to downtown.

        I don’t blame them. Sure, the bus is crowded, and slow. But the alternatives are worse. If your closest bus stop is on 24th and 57th or 24th and 61st, then you have to walk an extra five minutes, and wait a lot longer for your bus. The 40 comes every six minutes. Heading out of downtown, the 18 runs every half hour from 5:30 to 6:30, at which point it shuts down. Of course people take the 40.

        Running an express version of the 40 would help alleviate crowding, but so would simply running more 40s. The problem — in both cases — is that it would be expensive. The bus would not be much faster than the existing 40, and certainly not much faster than an existing 40 that goes on a stop diet. You would create bus bunching, which means that you pretty much have to provide the same level of service with the existing 40. It is expensive to run, because it isn’t that much faster. It gets stuck in the same traffic now, and if they add bus lanes, it could get stuck behind the other bus. That is a very expensive way to reduce crowding.

        It is quite likely that running regular extra 40s is actually cheaper. Let’s say you run 2 extra 40 buses. That means you go from 10 to 12 buses. That means buses every five minutes instead of six. Running 2 extra express buses would be cheaper (since it is a bit faster). The problem is, the express buses can’t be slipped in between the regular buses if you want to eliminate crowding (because they bunch). You can end up with worse crowding than when you started, because the time gap between buses went up (sometimes from 6 to 12 minutes). You have to run the express very frequently to eliminate crowding, whereas you can add as many regular 40s as you want (including just one) since you can easily space them out.

        In contrast, the big advantage of the 18 is that it bypasses much of the corridor, while providing a good portion of the riders with an alternative. It may not be a majority of those from 24th in Ballard — but even if it is a significant minority (say 10%, which is probably being conservative) that is still a lot less crowding on the 40. It reduces crowding on the 40, while providing a much needed improvement for those headed to downtown, along with the other areas along the way (including Belltown). If you are headed to Fremont or South Lake Union, you still take the 40, but if you are headed to downtown, it is better to take the 18.

  5. Two things come to mind:

    1) Fremont bridge both ways. Nothing realistic to be done about that.

    2) Waiting for buses to turn from Blanchard to Westlake northbound by Whole Foods. I’m hopeful something can be done here but not sure what.

    1. One thing that I don’t know why it isn’t done more often and think could help by the Whole Foods and all of Amazonia actually, is separating walk and car signals. Buses can’t turn from Blanchard to Westlake because everyone is lining up in the right lane to turn East in Denny. But often only two or so cars get to turn because of all the pedestrians, so the right lane never clears for buses to move into.

      1. Nobody should turn onto Denny from the south at Westlake. There are myriad alternate routes for people to do that. Westlake should become like its clone in San Francisco, Market Street, between Stewart and Denny: that is, transit, bikes and medallion cabs only.

        That would allow shorter cycles for Westlake, essentially just long enough for pedestrians to cross, giving more time for the diagonal streets making up the underlying grid.

      2. Westlake should become like its clone in San Francisco, Market Street, between Stewart and Denny: that is, transit, bikes and medallion cabs only.

        You don’t need to go that far. All you need is to make the right lane, between 9th and Denny transit only. Not BAT, but transit only. The left lane can then be used by cars. There are different approaches to take after that, but the simplest is to not allow any turns (left or right) from that left (general purpose) lane. Drivers have to go straight through Denny. They quickly learn that this is a terrible way to get eastbound on Denny, since it would require going way out of their way (

        This is not an ideal solution for the streetcar, but it fixes the problem for the 40 and C (south of Denny on Westlake). The C and 40 have a combined 24,000 riders, while that streetcar carries 5,000. At worst, if the streetcar is stuck, the riders would walk up and catch the far more frequent buses.

        That sort of small change (a bus lane one block long) is right up SDOT’s alley. They do that sort of thing all the time. It is also not without precedent for Westlake, since you can’t take a right on Thomas ( or Harrison ( Making that change should be fairly easy, while trying to make the entire street transit-only would be a much tougher challenge, with only a minor route (the streetcar) helped.

    2. > 2) Waiting for buses to turn from Blanchard to Westlake northbound by Whole Foods. I’m hopeful something can be done here but not sure what.

      Agreed 100%, this is a huge bottleneck. I submitted an idea for this via SDOT’s “your voice, your choice” program. Basically my thought is that we should just close the intersection to cars coming from 9th Ave that are trying to make a right turn onto northbound Westlake.

      You can upvote/comment on this idea here:, click “Other” ideas and then find the dot located at 9th and Westlake.

      1. It seems better to just address the root of the problem, regardless of how the cars get there. Some are coming from 9th, some are coming from the south (along Westlake), some are coming from Blanchard (along with the bus). The answer is to not allow them into the right lane of Westlake at all, north of Blanchard/9th. That is simply an extension of other parts of Westlake, which don’t allow cars in the right lane, and don’t allow right turns as a result.

    3. “1) Fremont bridge both ways. Nothing realistic to be done about that.”

      You can’t prevent the bridge from opening, but when it does, bus lanes could help a lot. Not only do bus lanes get the bus out of the line of cars, but they also allow the bus to reach the Fremont/34th stop and do all its loading and unloading while the bridge is up, so when the bridge goes down, it’s ready to go. Without bus lanes, the bus can’t even open its doors until the bridge goes down, the light turns green, and the cars in front of the bus move out of the way.

      1. I agree. The biggest problem with a drawbridge is not the time spent waiting for the bridge to open and close, it is the time spent waiting for the traffic to clear out. If you have bus lanes to the edge of the bridge, then the traffic is meaningless, and the bridge opening is not that big of a deal.

  6. I appreciate that they’re doing these mitigations, especially since I have recently moved from a city (Philadelphia) where they don’t bother doing anything like this. But I wish the approach wasn’t so piecemeal – one route at a time, only where slowdowns are identified.

    The Philadelphia approach (don’t do anything, really) gives the impression that bus riders are unworthy of any respect. The Seattle approach (expend a copious amount of effort to find the changes that give the most impact while inconveniencing drivers the least) gives the impression that bus riders are important, but not as important as drivers.

    The superior solution is to have every bus have bus-exclusive lanes along the entire route. This sounds crazy to most Americans, but it’s true: cars are always less important than buses. Not just during peak, not just when roads are congested, and not just when buses are unreliable, but always.

    1. +1 to Seattle > Philadelphia on these issues. I have been a regular transit user in both cities. Philly still has bus stops on every block and there seemed to be major apathy for transit riders among city leadership.

      1. I’m guessing Philly just has bigger problems. Seattle is blessed in that respect (even though we are starting to get some big city problems).

    2. Christopher, give Philadelphia a little bit of credit for this one.

      Started many years ago with original model cars that looked like streamlined PCC Streetcars, story is that designer rigged his own wind tunnel and watched the slip-stream shape his cigarette smoke.

      Sadly, don’t think I got to ride it. But gotta think how much skill a motorman needs to handle those accelerations and decelerations without standees being pasted to one end of the car or the other. Anybody with direct experience, please help save Philadelphia’s reputation.

      Mark Dublin

  7. There are some really horrendously long red lights throughout this route that ravage reliability. I drove the 40 every day last summer. My biggest headaches (and the worst time drains) were: crossing Aurora in both directions (but especially EB), crossing Greenwood Ave in both directions, the left turn from outbound 85th St to 15th Ave NW, the inbound left turn from 24th Ave onto Market, crossing beneath the Ballard Bridge, and of course the most notorious of all: southbound crossing Mercer.

    1. I don’t know if there is much that can be done about most of those intersections, but I’ve suggested that the 40 take a different path to Northgate. I think it should follow the route of the proposed new 61 ( That way the bus avoids making a turn at 85th and 15th. I wouldn’t send it all the way to Lake City, but just to Northgate.

      The would create a coverage gap for the northern part of the 40 (105th, Northgate Way, etc.). That could be covered two ways:

      1) Extend the D up to Northgate. I suggest this as part of my restructure proposal:

      2) Send the new 61 from Lake City to the layover at 8th and Holman Road (where the D lays over).

      At some point you want BAT lanes on 85th, but that is true no matter how the buses are restructured.

    2. > and of course the most notorious of all: southbound crossing Mercer.

      uggghh no worse way to start the day than waiting through multiple light cycles because someone in an SOV is blocking the box at Mercer for a bus full of 100 people.

  8. I’d just like to commend Metro for laying out Route 40,?esoecislly with the 13,000 weekday riders. Once Northgate Station opens in 2021, I expect even more riders will use it.

    I’m generally a fan of “C” shaped routes that begin and end at Link stations, and of routes that run between alternating residential, commercial and office areas.

Comments are closed.