Updated map

The Seattle Frequent Transit Map has been updated to reflect the Metro service changes taking effect this Saturday, February 15. The biggest service change is replacement of Route 358 with the RapidRide E Line that adds a new line to the evening frequent service map. There are minor design tweaks aiming to improve the clarity of route intersections. Thanks to Aleks for the suggestions.

Get the map here and enjoy! (8.5″ x 11″ 2 page PDF)

74 Replies to “Frequent Transit Map Updated”

  1. The timetable at left is largely blank; you forgot to mark the lines that aren’t frequent at all in the evenings or on weekends.

    1. Okay, I just checked on my IPhone, and it’s fine there. It looks like it’s a bug in Windows 8.1 PDF viewer, then. Sorry!

  2. Looks lovely, as it always has — thank you for your work on this.

    The evening map is no longer a total joke — it’s good to see the giant hole in north Seattle get filled in. Now if we can get evening frequent service down Delridge, unf*** the Eastlake/U-District service pattern, and squeeze out some hours somewhere to boost the Sunday evening 49 to frequent service, it would almost be a functional system to get around on in the evening.

    1. I’d also really like to see the 8 become full-time frequent. Instead of running the 43 every half hour after 8pm, just stop service entirely, and use the service hours for the 48 and 8.

      1. +1. With no evening 43, no one would have to walk more than a few extra blocks. UWMC riders could walk to Campus Parkway for the 49 or 70s, or to Montlake for the 255/545. Montlake/24th riders would have a frequent-service transfer from the 48 to the 8, Thomas St riders could walk 2 blocks to the 11, 12, or 10, E. John riders would still have the 10 or 49, and there’d still be the 47 every 30-40 mins for those on Olive/Bellevue.

      2. I think the structure of the 8 needs to be re-evaluated before a major injection of service subsidy. The section between Jackson and Madison is very little used (it has lots of riders on the bus, riding through from the RV to the Hill, but those will mostly evaporate in 2016).

        15-minutes evening headways on the Uptown-Capitol Hill and Mount Baker-Rainier Beach segments would be extremely desirable.

      3. Bruce: Yes, I was referring to the Uptown-Capitol Hill segment, which is the segment that would be losing service if the 43 were deleted at night and on Sunday. I also agree with you about the Mount Baker – Rainier Beach segment, but improving that segment using the 43’s service hours would be shifting resources between neighborhoods, which is something Metro generally tries to avoid.

      4. Wasn’t the 8 originally supposed to be full time frequent, until the continued existence of the 42 and then the budget issues changed that plan? If so, maybe Plan B passing, or a miracle in Olympia could make it a reality.

      5. You’re both right, even though for selfish reasons I’d hate to lose access to the 8 from my place at MLK/Cherry. It’s amazing that despite all the 8’s problems, it’s still faster to get to my downtown office on the 8 (25 minutes) than on the 3 (30 minutes). That’s why I just bike these days (15 minutes!)

      6. The 42’s hours were taken out of the 39, not the 8. The 39 (downtown – Seward Park) was extended to Othello Station to be a Link feeder and it was going to be half-hourly but it got reduced to 45 minutes. The 8 took over the Mt Baker – Rainier Beach segment from the 48. There was no plan I heard to make it full-time frequent.

      7. The 48 also has 30-minute headways all day Sunday, something that should probably be upgraded, although on weekday evenings, it does manage to maintain 15-minute headways until (I think) at least 10 PM.

        The real problem with the 48 is its lack of reliability.

      8. Well, as of next week’s, the 8’s frequent-service span will actually have shrunk , so that some hours presently used providing bus trips can be shifting over to useless schedule padding!

        Seriously. It’s in the service change announcement.

        It still won’t fix anything on Denny’s worst days. But at least the shoulder trips will now be routinely early and forced to drive with artificial lethargy, just as 40s and RRs and other increasingly padded Metro routes are. They will also screw up OneBusAway, which cannot compensate when Metro tells them two blocks = 5 minutes, leading to missed vehicles and rider ire.

        But most importantly, we now we have less frequency for the same money! Smart move, agency begging for cash at every turn!

        Seriously, though. This is getting ridiculous. We should not be presuming slow service to everywhere at all times. The padding festish needs to end.

      9. Bruce: Perhaps run the 8S as a rerouted version of the 27 or 14? Then the 8N can either run to Madison Park or just tie a knot somewhere in Madrona, Denny-Blaine, or the Madison Valley. Problem with that of course is you’ll get people riding the 8S instead of Link.

  3. Is the map meant to be bi-directional or only peak directional? The 21 southbound doesn’t have 15 min service until about 730am.

    1. I believe Oran’s map is based on Metro’s own frequency standards and reporting:


      But the intention is definitely for the map to be bidirectional.

      Technically, that page never actually says what time of day frequency *starts*. I haven’t looked through all the schedules in both directions, so I don’t know whether there are any other gaps like the one you describe for the 21.

      1. Right, the Metro page doesn’t have a start time but Oran’s map does. It says frequent service from 6am to 6pm.

        I’m not saying that the route shouldn’t be on the map but rather the start time should be taken off of Oran’s map as well to put it in line with Metro’s.

  4. The directional indicators on Roosevelt and 12th NE appear to be reversed. Roosevelt should point toward downtown (south or downward) and 12th NE away/north/upward.

  5. shouldn’t the 9/60 be included …
    9 on Broadway from Aloha to Madison (shared with 49 (from Roy), 60 (from Republican))
    60 on Broadway from Republican to Madison, and Madison to 9th Ave shared with (9, 49, then 12)

    or am I not understanding when and why you combine some routes like the 65/75

    1. It’s based on Metro’s own frequency standards and reporting:


      Some routes, like the 65/75, are scheduled (at least in theory) to have evenly-spaced headways. Others, like the 9/60, or the 28/40, or the 8/43 or 43/48, just coincidentally share a corridor. So they may provide 4 buses an hour, but that doesn’t mean there’s a bus every 15 minutes.

    2. They are not scheduled as a common corridor. There is up to a 20 minute gap between trips on the 9/60.

  6. Does anyone have a sense of how long it would take to walk from Rapid Ride E stop @ Aurora & Linn to Westlake to pick up the 40?

      1. Thanks. One wonders why RRE didn’t choose Crockett, rather than Linn for their stop. I’ll go digging for that answer. I’m sure it’s got something to do with stop spacing.

      2. Answered my own question. The RRE stops at Galer and at Linn. Galer and Crockett are too close together, and Galer picks up the walking overpass, and has stairs down to dexter and westlake. That’s a useful stop.

        In fact, I think that’s the better option for jumping over the the 40. Only 4 minutes down the hill with stairs.

      3. I’m struggling to come up with an actual pair origin/destination points where a transfer between the E-line and the 40 in Queen Anne would actually make sense. One would never do this to go to Ballard, as the transfer to the 44 is much more direct. SLU, the walking distance from Aurora and Denny does not seem enough to justify the overhead a transfer to eliminate it.

        I could sort of see taking the E-line from Shoreline to Queen Anne, then backtracking on the 40 to get to Fremont. However, even then, the numbers to justify it still don’t seem to add up. Walking distance from Aurora and 46th to Fremont and 34th is about 1 mile, estimated by Google at 24 minutes. By contrast, here is how long the trip would take with the connection: 3 minutes to travel on the E-bus from Aurora/46th to Aurora/Galer (by Google driving directions, assuming no intermediate stops and no traffic delay), 6 minutes to walk to the #40 stop at Westlake, 5 minutes to wait for the #40 bus (Somewhat generous, given that the #40 has 15-minute headways and is not super reliable. There is enough potential for delay downtown that even real-time info on OneBusAway back at the Fremont/46th stop would be next to worthless), 7 minutes to travel on the #40 bus to Fremont and 34th, plus another 3 minute to walk to the actual destination you could have walked directly to from the Aurora/46th bus stop without tagging the Fremont/34th bus stop along the way. Total time = 3+6+5+7+3 minutes = 24 minutes. Best case, the connection is no faster than walking all the way, yet probably considerably more stressful. Any delays waiting for the #40 to show up and the connection actually becomes demonstrably slower than walking.

  7. I’m with Gordon, but with the 372/522. I can see where the 372 and 522 combined create a 15 minute or less wait, but the last stop before downtown on the 522 is on 125th and Lake City Way, so that extension past 125th is misleading to me.

    1. I know that the map is geographically distorted, and there are a lot of places that don’t have much service but look well-covered.

      Having said that, it seems like there are two parts of the map that are undeniably underserved:

      – Northeast Seattle, between 65th and 125th
      – Magnolia

      I think we’ve spoken at length about how to fix Magnolia.

      As far as Northeast Seattle goes, if you look at a zoning map, it’s clear that the dominant corridor is Lake City Way. This makes me wonder if it would make sense to modify the 372 to become full-time frequent, and to run on Lake City Way between 65th (where it starts) and 125th. Going south, it could take Roosevelt/11th/12th to Campus Parkway, and continue through the UW Hub to become some other route(s). When North Link is finished, it would just terminate at Roosevelt Station.

      It seems like that one change could add a ton of geographical coverage to the frequent network. And because it would run along a fully commercial street, it wouldn’t draw any opposition from nearby homeowners.

      1. I seem to remember this being the 372’s route about a decade and a half ago. I’m guessing it was changed when University Village was revamped?

      2. Yeah Kacie, I remember riding with my dad on the old 372 that would go all the way down Lake City Way to 15th, then take that to 45th, then campus. I’d imagine it was moved to 25th serve that corridor (now that 15th is served by 65 & 66?) and to get out of the mess that was the left-turn onto 15th in the mornings.

        Aleks, Lake City Way runs in a topologically-separate region between 110th and 80th.. it’s hard for people in Maple Leaf to get down the hill, and people in Matthews Beach/Wedgwood to go over the many hills to isolated Lake City Way. The deviation you suggest would hardly serve any riders.

      3. Alper: Point taken. I didn’t remember that depression myself, but I guess I don’t spend all that much time on Lake City Way.

        It’s a shame that the main commercial strip in that area is so geographically isolated from the rest of the region, but that’s Seattle for you… :)

      4. It would also help if ST 522 would stop at LCW and 80th, especially with the poor evening/weekend service of the 72/73 routes.

      5. LCW is a transportation corridor between the hills. Where the hills are steep it’s hard to access from residential areas on the hill; it shares this feature with many Seattle transportation corridors, from the Burke-Gilman Trail to MLK. Between 110th and 90th or so it’s pretty hard to access, from the west especially. The buildable depth of this corridor is as limited by terrain as access is — very limited between 110th and 90th and less so elsewhere.

        Other N-S corridors: 35th Ave (from 65th all the way to Lake City) is a moderately important commercial corridor, though not unbroken. There’s some momentum for planning growth along 35th, which seems like a mixture of developers getting out in front on messaging and some, “Not one bit of growth a half-block off 35th,” type stuff. On 25th (the 372’s current route) there’s not much between 55th and LCW but the 372 picks up lots of people south of 55th just going to campus.

        I think routes doing what the 65 and 75 do will probably be around as long as there are buses in Seattle, even if they don’t come frequently. Similarly, some route will have to take people between 25th and campus south of 55th. As for the rest… there are choices.

    1. Yeah, I have better transit access in Kenmore than I did when I lived in Wedgwood. Though Northgate Link could potentially make Wedgwood a lot better if they run frequent buses to Roosevelt station.

      1. It’s not like Wedgwood is a black hole for transit, it’s just that none of the many routes through there are coordinated or consolidated into frequent corridors. It’s definitely true that the critical moment to reform this will be a restructure around the time Roosevelt Station opens.

      2. 35th is ripe for a streetcar. Can you imagine something that starts at 85th/95th and running all the way down to U-Village and onto the University Stadium Station? :D

      3. @Alper,

        This is satire, right? 35th is a four (narrow) lane wide street with frequent parking along both sides. A streetcar would be dead meat to left turners and would require middle of the street boarding. San Francisco gets away with that by having a stop sign at nearly every LRT stop, but you couldn’t do that in Seattle.

      4. Wedgwood isn’t really that bad for transit, but getting downtown at non-peak hours was pretty slow. Your options were the 71 via the U-District, or the 65 and transfer at the U-District.

        If I were in charge of re-routing the buses for Roosevelt station, I would probably truncate the 71 at the station and keep the 65 more or less on its current course.

        I don’t see a streetcar for Wedgwood anytime soon. Buses will be fine for the next few decades.

  8. I still really bristle at how this map condenses the distances in South Seattle so much more than in North Seattle. It suggests that South Seattle is well-served when there are long distances and slow travel speeds that plague the area. It’s clearly a North Seattle skewed view of the world.

    1. The goal of the map is to illustrate where service *exists*, not where it *doesn’t*. As a guide for getting around the city, it’s much more legible than a geographically-accurate map would be. Pretty much every useful transit map magnifies areas with more service and shrinks areas with less; that’s just the nature of mapping.

      As a political statement or rallying cry, I agree with you that a more geographically accurate map would be useful. Really, what you want is a heat map, that shows density adjusted by walkshed to frequent transit service. You’d see a lot of green in places that are either not very dense or that have a lot of transit; you’d see some yellow hotspots in Magnolia and Wedgwood and West Seattle; and you’d see a big chunk of red in South Seattle.

      1. I do like the orientation of the map to focus on only frequent routes. It also does a good job showing some areas that have weak service (Wedgewood) as well as other areas that don’t connect well (West Seattle). Is it possible to have a map to show speeds on these routes? I’d think that would be quite revealing. It seems that we are evolving to have two different speeds of transit service — local and limited stop/RapidRide/Link.

  9. I know it’s a crowded intersection (on the map what with the 40 and 48) and perhaps nowhere to put it, but the Rapid Ride D is missing the (important) stop at NW 85th Street.

  10. Has there been any consideration of extending the E line further south to Georgetown? That would be huge for that neighborhood, and it seems like a natural extension.

    1. The main problem with this is that the 358 is already one of the longer routes that Metro operates, and it’s not particularly reliable, either. So Metro doesn’t want to extend it any further south than downtown. Even in David Lawson’s “Santa” network, his version of the 58 terminates in downtown.

      While I agree that Georgetown could use better transit, I don’t think it’s first in line for the RapidRide treatment. The combined ridership on the 131/132 is 6,000; high, but not earth-shattering, especially since the common segment is probably lower.

      Other bad candidates for RapidRide include the following:

      – Any route that will soon be replaced by Link, such as the 41 or the U-District expresses, or (arguably) the 40. I’m also including routes whose ridership may be reduced by rail, such as the 60 (FHSC).
      – Trolleybuses, and routes that *should* be trolleybuses. Trolleybuses have unique operational and capital challenges, and it would be expensive to retarget the RapidRide program for those routes, with little benefit. I would like to see BRT elements on the trolley network, but I don’t think RapidRide is appropriate.
      – Any route that should really be part of Sound Transit’s network, such as the 150, 255, or 271.

      Based on ridership, the next RapidRide line (G) should be the 120. With a ridership of 8,600, it’s one of the highest-performing non-trolleys (or trolley potentials) in the system. It would also be able to reuse some of the transit infrastructure that’s already been built for the C Line, or that might be built in the future.

      I think the 5 (H) is also a great candidate, especially if it were rerouted through Fremont. Just by itself, the 5 has 8,000 daily riders. Picking up Fremont would probably give it another 2,000 or 3,000 riders. While it’s geographically close to the D and the E, it serves an important corridor with a history of strong performance.

      Georgetown seems like a good #3, though. Maybe the “I Line”? ;)

      1. How do you think the D, E, & H (5) lines should be affected when Link Northgate opens?

        I’d like to see some connections from Greenwood, Phinney Ridge, etc. to Northgate.

      2. The D should definitely go to Northgate, in lieu of the 40. There are complex operational reasons why it doesn’t work that way now, but I still think it should and will happen eventually.

        I think the 5 and 358 are more useful as additional N-S corridors than as connections to Northgate. The 5 used to go to Northgate half the time, and ridership improved dramatically when Metro straightened out the route. The 40 connects upper Greenwood to Northgate today. If the 5 were a RapidRide route, and the D went to Northgate, then the connection would be pretty seamless.

        Ultimately, I’d like to see Ballard Link with stops in Fremont and upper Greenwood (105th). That will provide a huge connectivity improvement.

      3. In addition, the 48, or some variation in the future, connects Greenwood with Link light rail at Roosevelt Station. From there it is a one stop ride to Northgate on the train. Yeah – you backtrack about 20 blocks, but you can blame I-5 for that.

      4. I sure hope it is easier to get from Greenwood to Northgate Link than taking two routes. It currently takes me 47+ minutes to get to Stadium Station and I’d like to cash in on the future 22* minute Link ride from Northgate to Stadium Station without wasting my time savings in routes just to get to Northgate.

        + Scheduled for 47 minutes, realistically takes longer
        * Based on 14 minute ST estimate from Northgate to Downtown (assumed to be Westgate) and current 8 minutes from Westlake to Stadium Station

    2. RapidRide is significantly more expensive than just making a route frequent. You have to order a limited-run set of buses and paint them red, install a fiber-optic wire and ORCA readers along the route, and convince SDOT to upgrade the street and set up signal priority. So it’s an open question whether there will ever be any more RapidRide routes. The federal program that subsidized them is now going in two different directions. One is higher toward “real” BRT. The other is lower toward incremental improvements (no distinct brand). So Metro will probably aim toward the lower level too; e.g., making the 120 and 5 full-time frequent without rebranding them. Street improvements for the 120 are already in the works.

      1. I didn’t know that about the federal program. Thanks for the info. I think the separate brand is kind of silly, so I’m not sad to hear it. I still think the 120 and 5 should be next in line for improvements. And I still hope that Metro continues to steadily replace frequent couplets with frequent single routes (e.g. 15/18 with D, 3/4 with 3, 2N/13 with 23, 26/28 with 16, 66/67/71/72/73 with 73, 54/55 with C, 230/253 with B, 5-Shoreline/5-Northgate with 5).

      2. I wish KCM would take up “real” BRT since Link to most of these areas is so far out. I’d also like to see them ditch the free WiFi in favor of faster and more frequent routes and enforce backdoor unloading so we can gain 5 minutes back on ever busy route. Takes me 47 minutes on the 5 to get to stadium station from Greenwood but only 15 in a car… how am I ever supposed to convince people to take public transit?

  11. Does the 24/33 from Magnolia (28th/Thorndyke & Blaine) to Downtown Seattle count? I know it’s not listed on Metro’s list, but the combined service seems to be every 15 minutes at least on weekdays (although there may be some exceptions that I haven’t seen yet).

    Let’s hope that when University Link opens, Metro actually restructures service so that there is more frequent service in Northeast Seattle. Even now though, it seems like Metro could just add 2 stops to the 372 and stagger the 68 and 372 appropriately so that an additional frequent corridor on 25th Ave NE until 75th St is created for no cost.

    1. So, if you’ve got two buses that each come every half hour, there are a huge number of different possible service patterns. At one extreme, the buses can come at precisely the same time. At another extreme, the buses can come precisely every 15 minutes at alternating intervals. Somewhere in between, you’ll have spacing like 25-5, or 20-10, or something like that.

      Let’s round to the nearest minute. Of the 16 different possibilities, all but one will lead to at least two periods every hour during which a bus will take longer than 15 minutes to arrive. Only if the buses are perfectly spaced — one possibility out of 16 — will you be guaranteed to have 15-minute headways. Therefore, if two bus routes aren’t designed to be perfectly spaced, the balance of probability suggests that they won’t be.

      In fact, a quick look at the 24/33 schedules suggests that the service pattern is 20-10 or 21-9 for most of the day.

      Honestly, Metro’s network has this problem in spades, especially evenings and Sundays. There are almost always four buses per hour on John/Thomas, on 23rd Ave, on Pike/Pine, on Madison, and on many other important corridors. And yet the poor spacing means that, after 8 pm, you generally have to wait at least 20-25 minutes for a bus. Magnolia is just particularly unfortunate in that this problem persists 24/7.

      1. For Magnolia, it seems that in the midday where both the 24 and 33 run on clock-face schedules, they are actually exactly 15 minutes apart. From 28th and Blaine to Downtown, the 24 leaves at about :00 and :30 while the 33 leaves at about :15 and :45. There do seem to be some issues in peak hour where clockface patterns break down: for example, buses depart Magnolia at 2:33, 2:44, 2:50, 3:18, 3:23, 3:46, etc. where there is suddenly a 28-minute (!) gap due to crappy scheduling. Metro, this is really not that difficult to fix…

      2. I think I was looking at the outbound schedules. Regardless, I agree with you that this would be an easy and almost cost-free way to improve service.

      3. Awhile back I’d looked at the schedules for the 8, 43, and 48 to see if it would work to shift them around to even out the spacing on John/Thomas and 23rd/24th. At the time moving the 43 by a few minutes was all it would take, but then I discovered that the 43 and 49 are scheduled to have even spacing departing from downtown (and roughly even from the U District). So much for that.

      4. David: You’ve touched on the biggest problem with the 43, which is that it’s the single most redundant route in Metro’s entire network. Every stop and trip and corridor it serves is also served by another route, or possibly a combination of two routes. You’d like to coordinate the 43 with some combination of the 8, 10, 11, 47, 48, and 49, but obviously that’s impossible, so the best Metro can do is to pick one and hope for the best.

        In general, I’d like to kill the 43 and use its service hours for other routes. But I think the argument is a much easier sell during the time periods when the 43 already runs at half-hour frequency. If Metro can only afford to run four buses per hour on John/Thomas and 23rd, the least they can do is make sure they’re evenly spaced.

      5. The 43 may be grid-incorrect but it’s one of the post productive routes on Capitol Hill and connects a wide variety of useful trip pairs that no other route does. It’s also an existing trolley route that doesn’t need any new wire. Some of the proposed transfers have such short segments they get silly. The 8 is not close enough to downtown to substitute for the 43 so you’d have to transfer to the 47. If your destination is on 23rd/24th you have to transfer at 23rd & John for just another mile or half-mile. Meanwhile the majority of trips are along UW-Montlake-23rd-John-Bellevue, which is exactly where the 43 already goes.

        However, having said that, I’m intrigued by Aleks’ idea of making it a daytime-only route. That would be more palpable than deleting it entirely, and the vast majority of trips I’m describing are in the daytime. Having to switch to the 8+48 or 49 in the evenings and walk a few blocks more would be a reasonable tradeoff. (Although women travelling alone may think differently.)

        This then gets into the 49, which is currently the designated frequent route for Capitol Hill. I’ve thought significantly about Capitol Hill restructures, and they all come back to the 43 and 49 being the most grid-incorrect routes but paradoxically the highest-ridership. I would rather split the 49 than the 43. A more frequent 10 could take some of the 49’s riders, and a one-seat ride to 10th E is less useful than a one-seat ride to east John and 23rd/24th. But Metro has been positioning the 49 as the main route for Capitol Hill.

      6. I don’t see much point in trying to space the 43 and 49 to have even spacing departing from downtown, as there are very few trips where it would make sense to treat the two routes as a combined corridor. For example:

        Downtown to Capitol Hill: There are tons of buses going down this corridor, and the 43 and 49 are only two of them. (Add the 10, 11, 12, and 47). As long as the combined frequency of the 10, 11, 12, 43, and 49 is decent, the combined frequency of the 43 and 49 alone doesn’t really matter.

        Downtown to U-district: Neither the 43 nor the 49 is an efficient way to make this trip. The 71/72/73, even running the local route, are usually faster than this, except in extreme situations (like when they’re overcrowded and likely to pass you up).

        Capitol Hill to U-district: If you look at the map, the 43 and 49 take different routes through Capitol Hill. They have one intersection in common, Broadway and John. In practice, both routes are frequent enough individually that routing decisions tend to be based on which route comes the closest, rather than which route is coming first. In any case, because the two buses take such different routes (not to mention the 49’s thru route with the 7), any attempt to coordinate the 43 and 49 for a frequent service corridor would be complete joke anyway.

Comments are closed.