San Francisco Planners Proud of Long List of Road Diets

Valencia_small.jpgValencia Street bicycle lane and stencil. Photo:KayVee.INC

Although there is no record book for the cities with the most traffic-calmed streets, San Francisco planners believe their city has the most road diets, or roads that have had auto lanes narrowed or removed to calm traffic speeds and provide room for other modes of travel.

The San Francisco Municipal Transportation Agency (MTA) has instituted 34 road diets since the city’s Transit First policy was adopted in 1973, a number that is expected to grow once the bicycle injunction is lifted and new bike lanes are striped.

MTA traffic engineer Mike Salaberry said the city had a number of roads where lane width was excessive and designed only for cars. Road diets, he said, were primarily "to improve how the road works for a wider range of users."

"Older roadway designs were not efficient for using space," said Salaberry, who pointed to numerous streets where wide travel lanes led to long crossing distances for pedestrians and no dedicated space for cyclists. "Given that space is at a premium, we had to find creative ways to allocating the space we do have."

In some cases, making space for pedestrians and cyclists is as simple as painting lines on the street and reducing the opportunity for drivers to reach high speeds. Walk San Francisco’s Manish Champsee hailed the MTA’s track record on road diets and said they were doubly beneficial for pedestrians. Slower speeds mean less collisions and less-severe collisions when they do happen, which dramatically increases a pedestrian’s chance of survival. Another benefit comes from visibility: On a four-lane road reduced to two lanes, pedestrians don’t have to worry about the danger of the car in the lane further from the curb not stopping for pedestrians at un-signalized crosswalks.

Champsee pointed to the Valencia Street road diet as a success story that has become a national model for traffic engineers. Valencia was a four lane road until March 1999, when the former Department of Parking and Traffic (the DPT is now part of the MTA) re-striped the street to its current configuration, with two travel lanes, a center median with left-hand turn bays, and bike lanes.

Despite DPT concerns at the time that bicycle collisions would increase, in the year after the lanes were striped, total bicycle collisions and pedestrian collisions declined, and there was a 144 percent increase in bicycle use. Though some vehicles diverted to other streets, vehicle volume was only slightly lower, and it didn’t all move to Guerrero Street, where traffic engineers expected it to go. Fears of negative economic impact were also unfounded. As an SF State study noted subsequently, only 6 percent of merchants surveyed after the bike lanes were installed had negative feedback.

"There is really nothing bad you can say about what happened on Valencia Street," said Champsee, who added that he was at a recent Federal Highway Administration forum on street safety and Valencia was an example the federal agency used to highlight a successful road diet.

Despite the proven positive impacts of road diets in San Francisco, the hurdles to further changes are significant. As the protracted battle over the Bike Plan illustrates, road diets have to undergo significant environmental review and if they impact on traffic flow, they either get scrapped or have to receive an exemption from the city.

"As always, we’re up against some of the perverse unintended byproducts of the California Environmental Quality Act (CEQA)," said the San Francisco Bicycle Coalition’s Andy Thornley. Despite San Francisco’s long-standing Transit First policy and the unquestionable benefit of slowing vehicle speeds, projects can’t be expedited faster than they currently are.

"Over and over again, these projects illustrate what’s counter-intuitive to some folks," explained Thornley. "If you take four lanes and make them three, it seems that you’re losing something. In fact, by providing a left turn lane, the throughput of the street may remain the same or improve, while the safety improves greatly."

The next big road diet project is the redesign of Cesar Chavez Street from Guerrero Street to the Highway 101 "Hairball" interchange. The city will remove one lane of traffic in each direction, widen the median and install left-turn lanes, and add bicycle lanes. The project will not only represent a new level of city agency vision and cooperation, but if successful, could set a new standard for livable streets in San Francisco.

Thornley said Cesar Chavez and other similar projects like 25th Avenue in the Richmond would further establish the benefit of traffic calming and multi-modal planning. "In spite of what drivers may think, it will make their trip safer and it might make it faster."

List of San Francisco Streets with Road Diets

Courtesy: MTA

  • Arguello from Pacific to Fulton
  • Valencia from Market to Tiffany
  • Polk from Turk to Vallejo
  • Mansell from San Bruno to University
  • Harrison from 11th to 22nd
  • San Jose Avenue, southbound, from Randall to Arlington
  • San Jose Avenue/Guerrero from Randall to Cesar Chavez
  • Dewey Boulevard from Laguna Honda to Taraval
  • Post Street from Presidio to Steiner
  • Turk Street from Arguello to Masonic
  • Golden Gate Avenue from Masonic to Broderick
  • Fulton Street from Baker to Webster
  • Fell Street westbound from Scott to Baker (Laguna to Scott PM tow-away later removed also)
  • JFK Drive westbound from Shrader to Conservatory Drive East
  • Duboce from Buchanan to Church (conversion from roadway to bikes only)
  • Fourteenth Street from Dolores to Guerrero
  • Howard Street from Fremont to 11th
  • Seventh Street from Townsend to 16th
  • Oakdale from Phelps to Industrial, Bayshore to Industrial
  • Battery Street from The Embarcadero to Broadway
  • San Bruno from Mansell to Campbell
  • Folsom Street from 14th to 11th
  • Market from 8th to Van Ness
  • Potrero from 17th to 25th
  • Lake St from Arguello to 3rd Ave
  • 14th Street from Market to Dolores
  • Alemany from San Jose to Rouseau
  • Clipper St from Douglass to Diamond Heights
  • 25th Avenue from Fulton to Lake
  • 7th Avenue from Lawton to Judah
  • Oak Street from Divisadero to Laguna (removal of AM tow-away lane)
  • Scott Street from Oak to Fell St
  • Gough St northbound from McCoppin to Market
  • patrick

    Now lets starting putting some roundabouts in the middle of intersections like Berkeley has done. They not only calm traffic, they add green space and reduce the amount of run-off that enters the bay.,_Berkeley.JPG

  • JohnB


    I like roundabouts and I think they can work well. Clearly they do in Europe where they are everywhere.

    And it’s great not having to stop at every block if the other roads are clear, unlike the 4-Way stops in SF.

    But first people here have to learn how to navigate them. They tried a couple around Duboce/Lower Haight a few years ago and the locals rioted, after which they were soon removed.

    Once burned; twice shy.

  • Another for the list: Fillmore, which seems to have had two southbound lanes in 1985.

  • pde

    Patrick, a transport engineer friend of mine from Melbourne, which has lots of roundabouts, told me that they are comparatively dangerous for cyclists. There are other kinds of green space measures that are better, such as blocking off a street entirely in a few places, with bike lanes through the barriers. See eg,-95.677068&sspn=34.122306,56.513672&ie=UTF8&hq=&hnear=Melbourne+Victoria,+Australia&ll=-37.788929,144.974427&spn=0.016618,0.027595&z=15&layer=c&cbll=-37.789087,144.974399&cbp=12,212.67,,0,1.92


    These prevent through traffic, but still preserve accessibility for local traffic.

  • patrick

    Webster between Hayes & Bush could definitely use a diet.

  • patrick

    pde, that’s interesting, but I’ve never heard anything like that before. Why would they be dangerous for cyclists? I’ve biked around them in Berkeley and they didn’t seem at all unsafe to me.

  • Patrick, those ones in Berkeley I’ve biked are on quiet residential streets. Already slower and more sparse traffic might help in making it feel safer. I’ve never had trouble on the ones in Berkeley either.

  • patrick

    yeah, I did a little research on it, and from what I could find there are some issues, but they are mainly with high traffic roundabouts where there is a bike lane on the outside, as drivers are not looking for vehicles in that location. The modern recommendation for roundabouts is to have no bike lane, and have cyclists merge in with drivers.

    The type roundabouts I’m referring to do not have the above problem, as I mean for low traffic intersections, like they do in Berkeley.

  • david vartanoff

    Having biked in Berkeley, two issues arise. First, as the plantings grow drivers and cyclists vision of each other is impaired. Secondly, despite signage indicating correct routing, many times I have seen cars make a clockwise left 90 rather than the correct counter clock 270 followed by a right 90.

  • Jeremiah

    Some of the roundabouts they use in Germany and other parts of the continent do work well for cars and bikes, but they have the bikes on a fully separate lane that crosses next to the crosswalk, which in proper roundabout design is separated (set back) from the main circular part of the intersection by perhaps 10′, and often on raised sidewalk platforms. .

    This requires the cyclists to slow and behave a little more like pedestrians at these intersections, but they don’t mix with the pedestrians or the cars. And the car movements are slowed considerably by the design of the intersection, and the cyclists are in full view of the drivers.

  • Nick

    The City is limited to what they can install based upon funding. In other words, there’s not going to be a new traffic circle anywhere in San Francisco in the next 10 years.

    They cost something like $100,000 each. Some neighborhoods have homeowners associations where dues are collected to pay for maintenace and things such as this. One example that comes to mind is that Traffic Circle/ Water Fountain in the intersection of Saint Francis and Santa Ana.

    If I could borrow a photo (fountain in the distance):

  • Joel

    I can think of so many places in SF that still need these road diets, esp. in the southern part of the city. Take Ocean Ave. west of 19th Ave. This is a 75ft. wide expanse of asphalt begging for a median or even a bike lane (so simple!). There are four schools a block away along the narrower Eucalyptus Dr. that would like to see these improvements. Students and teachers at Lowell High want these kinds of improvements.

  • John C.

    Imagine our beloved Fell street after a road diet. One traffic lane in each direction, each with a bike lane. Ahh, so nice.

  • patrick

    Nick, it’s not a question of funding, but of priority and desire to do it. There’s plenty of money being spent on roads. I’m not saying it’s going to be spent on roundabouts, but if we all wanted it done, it would get done.

  • i can honestly say that Cesar Chavez is the worst road I’ve ever ridden my bike on — and it’s not even close.

    speaking of CEQA — i think that should be our top statewide legislative agenda item — forget everything else — let’s modify CEQA.

  • Plannet

    Don’t forget the greatest SF road diets of all: the Embarcadero Freeway and the Central Freeway!

    Next on the list for conversion into a lovely boulevard: I-280

  • ZA

    Sadly, there was a nasty accident on Cesar Chavez just two days ago, in the afternoon commute, on the approach to that ‘Hairball.’ I believe 3 cars were involved, one of them intruding on a house…an all too common occurance on that road.

  • There are differences among Berkeley bikers about roundabouts. Some of us back them because they slow traffic, and others dislike the conflicts that they cause between bikes and cars.

    The problem is that cars do not know that they should merge into one lane with bikes at the roundabout. If both a bike and a car are approaching a roundabout, the car will usually veer to the right to go around the roundabout, without realizing that this endangers the bicyclist.

    The roundabouts work pretty well for bicyclists who are aggressive enough to take the lane while approaching them. They don’t work as well for less aggressive bicyclists.

    This could probably be helped by signage telling cars to slow to 15 mph while approaching the roundabout and to merge into one lane with bikes while going around it.

  • patrick

    I’m not sure I understand the situation you are describing, is this a multiple lane roundabout? I’m not talking about multi-lane roundabouts, just putting them in the middle of intersections where there is only 1 lane of traffic in each direction, like in the picture I linked to above (although I guess it’s hard to tell that’s the case from the image).

    Well, I guess either way I agree roundabouts are not perfect, I just prefer them to the intersections as they exist today.

  • Mike

    Nick, not sure where you’re getting your information but there is funding and plans for traffic circles in the next couple years in San Francisco. People are generally leary of them so we’re going to have to be careful and work closely with the immediate community and stakeholders.

  • tea

    I’m completely in agreement with Joel when he says, “I can think of so many places in SF that still need these road diets, esp. in the southern part of the city.”

    The southern part of the city is chock full with high-speed residential freeways (Alemany, Ocean, San Jose, Geneva etc etc). Very little is done or even discussed about that, yet it affects just as many people as Cesar Chavez and Valencia.

  • There are little landscaped traffic circles all over residential neighborhoods in Seattle. Very calming and civilized. The beauty of a roundabout or traffic circle is that cars (or bikes) don’t have to come to a complete stop if there is no other traffic in the circle, but they should slow *way* down (less than 10mph) before entering.

    Here’s a nice write-up of Seattle’s traffic circle program. Where implemented, the circles have dramatically dropped accident rates (by 94%!) and injury rates from 153 per year down to 1. (Hard to believe, but these truly are the results.) These circles cost $3000 – $6000 to build.

  • Just wanted to give my experience with roundabouts in Parkmerced commuting to SF State for 3 semesters by bike and sometimes scooter – they were a nightmare for me, but I think it may be that those are just designed with the lanes too large. On my bike I always felt like an ant holding up the line, and I once fell off my scooter when someone cut me off trying to jump in in front of me. I think there’s a lot of visibility problems that direct drivers to look in the wrong places. They were also terrifying to walk on, feeling like a car could just stop turning and go straight into you, particularly in the crosswalk. Not that I’m against roundabouts, just the way those ones have been designed.

  • JohnB

    Charles and others,

    Note that roundabouts only slow or “calm” traffic speeds if they are placed at intersections that previously had no stop light or sign.

    Ironically they are often introduced in place of stop signs or lights to actually speed up traffic flows, since a stop is no longer required if your exist path is clear, and drivers can now roll through” a roundabout in the same way as many cyclists “roll through” stop signs and lights.

  • patrick

    JohnB, that’s not true. They can calm traffic by making it impossible to simply run a stop sign, they also prevent a driver from gunning it immediately past the stop sign. A driver is physically required to move at a slower pace through the entire intersection when there is a roundabout.

    As Aaron and others have pointed out the devil’s in the details. While the main issues with roundabouts (as far as I’ve been able to find) have been solved from a technical perspective, it still requires competent design to make sure they are implemented correctly.

    Unfortunately all the roundabouts I’ve seen in SF have been poorly designed. The one Aaron linked to has a tree in the middle obstructing visibility, and the other one I know of (8th & Townsend) is extremely poorly designed and has an excessively wide lane.

    Traffic “calming” is not necessarily about slowing vehicles, it’s about making a safer environment for all, and roundabouts can actually be used to increase throughput, yet provide improved safety at the same time, but it does depend on the situation. They are not a perfect solution for every situation, but they certainly have some significant benefits.

  • patrick

    wow taomom, that was a great link! Seattle’s program seems very well thought out and implemented. I recommend anybody who has questions regarding roundabouts (particularly the ones I’m talking about) take a look at the link taomom posted above.

  • Joel

    @tea – For San Jose Ave, there is talk about creating intersections at the part close to I-280, which could have some downsides, but it would slow traffic and rejoin the community. It’s part of the Glen Park Community Plan ( but I think it needs more momentum.

  • JohnB


    If we want traffic-calming, then there are various ways to do that. No problem.

    I was responding specifically to Charles’s assumption that that is the primary purpose of a roundabout or circle. It isn’t. In Europe, where they are used extensively, they are designed to actually increase the average speed and throughput at an intersection. Essentially it lets the constantly-changing flow of traffic dictate priorities for proceeding, thereby increasing the number of vehicles per hour.

    Now, having said that, if traffic calming is your SOLE purpose, then you can design a circle which will slow things down. And in fact I’ve seen some that also have stop lights. But that’s not the original purpose of the roundabout – quite the opposite as I have said.

    And yes, a driver can’t blow through a roundabout the way he might blow through a red light. But then frankly I see cyclists doing that more than cars anyway. The real point is that you don’t have to stop at a roundabout if nothing else is approaching the way you would at a light or sign.

    So if we are going to put in circles, we need to be clear exactly WHY we are doing that, and then design them with that in mind. But the experimental ones in the Lower Haight were a failure on every measure.

    And Americans don’t seem to know how to negotiate them so expect some accidents at least in the early stages.

  • patrick

    JohnB, I’m not really concerned about the original purpose of them. Please read taomom’s link. It describes the Seattle program: A U.S. city, highly car dependant, and they saw a huge drop in accidents and injuries in the first year. They’ve installed about 600 hundred over the last 2 decades and they are very popular, they have far more requests for them than they are able in install.

    sure roundabouts can be used to improve traffic flow, and I think that’s a good purpose for them too if it also improves safety.

  • Mike

    Looks like we need a streetsblog article on traffic circles and roundabouts! The two designs seem similar but have some key differences.

    A neighborhood traffic circle is what Seattle uses extensively and what taomom refers to. These are smaller than roundabouts and are generally used to slow traffic through intersections where motorists would otherwise be able to go straight through. They help reduce long sight lines down a road that encourage higher speeds and also serve as an opportunity to landscape and beautify an intersection. The right-of-way rule is that the person on right has the right of way if you arrive at the same time.

    A roundabout is larger – sometimes with multiple lanes – and has the right of way rule where you must yield to traffic already within (ie to your left) the roundabout. These are often used in higher volume intersections to maintain traffic flow with much fewer conflict points.

    Both types of measures work very well to reduce crashes and allow slower, steadier movements through an intersection. And both measures are generally misunderstood and feared in the US, despite their record of success in many parts of the world. We’re slowly getting it though, with two examples being the succes in Seattle with circles and even Caltrans using roundabouts (like in Truckee near I-80) more commonly.

    The Page Street experience (with the smaller traffic circles) did not go well for a number of reasons that did not necessarily have to do with the usefulness of circles themselves. Crashes, for instance, dropped when those circles were put in as a trial.

    Anyway, seems like we can use some more discussion about these things.

  • patrick

    Thanks for clarifying the difference Mike!

    I was referring to traffic circles in my comments above.

  • JohnB


    A Traffic circle will still speed traffic up if it is used in place of a sign or light that would otherwise stop traffic. I thought that was very clear.

    It only slows down traffic at an intersection which would otherwise have neither a light nor a sign


    I like your distinction between circles (small) and roundabouts (large). In Europe there are also super-large roundabouts, often the size of a city block, which have smaller roundabouts around them. This is sometimes called a Gyratory System.

    In all three cases though, the intent is to speed up traffic flows. If you wan to slow traffic down you’d instead use street narrowing, chicanes, bollards, speed bumps or other means. A circle of any type per se doesn’t calm traffic except as an obstruction

  • The traffic circles that have been so extremely successful and popular in Seattle take up enough of the intersection that they force cars to go slowly. I have witnessed them in action, and they really and truly do slow traffic, and more pleasantly than a stop sign every block. (Race, stop. Race, stop. Or not stop, as is often the case with San Francisco drivers.) They have the added benefit of giving the sense that you are driving through a neighborhood rather than on a street that happens to have people living on it.

    The time may come in San Francisco when we wonder why we’ve sacrificed beauty, community, and safety in our city all in order to get from point A to point B five minutes quicker.

  • JohnB


    From what you appear to be saying, the Seattle “circles” function more like deliberate obstacles to vehicles than to aid traffic flow.

    That’s fine if that is the sole purpose of them. But that makes them more like the “islands” they use in Berkeley specifically to calm traffic. And note that they are used on lightly-used residential streets there. it’s not going to work where, say, Fell crosses Divisidero.

    If we can focus major traffic flows on the main arteries e.g. by roundabouts, then that gives more opportunities to place floral islands, barriers and chicanes on the quiet residential streets to enable the pleasant and scenic ride you crave.

    Let’s try and help ALL road users, rather than make it a battle between them.

  • patrick

    JohnB, “A Traffic circle will still speed traffic up if it is used in place of a sign or light that would otherwise stop traffic. I thought that was very clear.”

    why do you keep talking about slowing or speeding traffic, and in particular shy are you addressing the comment to me? I already said above I see improved traffic flow a bonus if it also increases safety.

    Also, why don’t you actually read the link posted by taomom, many of your comments questions are addressed in it.

    In fact, I’ll post it again so you don’t have to scroll back to find it:

  • JohnB


    I’m familiar with the Seattle “islands” which, as I noted above, are similar to the ones in Berkeley.

    Whether we think roundabouts are primarily for safety or throughput is a matter we can debate and differ on. But I’d extend Mike’s taxonomy of terms to define four variations:

    1) Mega-roundabouts AKA Gyratory systems, often consisting of more than one circles, to handle intersections of multiple major highways

    2) Regular roundabouts, to improve traffic flows at busy intersections

    3) Traffic circles, small scale, more geared towards flexible priorities on less-used streets or where a minor street intersects a busy street

    4) Islands, such as in Seattle and Berkeley, where safety and traffic-calming are more the aim.

    So with all that choice, I feel sure there is a roundabout there for everyone. And I’m equally sure someone here will still not be happy. But remember that the Arc de Triomphe in Paris is a traffic circle, writ grand. It’s not a new idea.

  • patrick

    You may be familiar with the Seattle traffic circles, but the reason I ask you to read the link is that you are making statements that are contradicted by it.

    Perhaps you have other information that you are basing your statements on. If so, please share it. But making statements without support is not helpful to the discussion.

  • JohnB


    I’m sorry but I don’t have to agree with every point made in a cited article. I hope that’s OK with you.

    I have many years experience of navigating different roundabout systems in Europe, and I have shared my resultant experiences here since I know most in the U.S. have no idea how they work and what they are for.

    I believe that the taxonomy I provided above provides all the possibilities that folks here have expressed an interest in, and then some. Take your pick from them. That’s the beauty of having a choice.

  • patrick

    you can disagree with whatever you want, but if you are going to disagree with something that has been studied in detail you should provide support for your argument.

    I’m not talking about your taxonomy, I’m talking about statements like:

    “From what you appear to be saying, the Seattle “circles” function more like deliberate obstacles to vehicles than to aid traffic flow.”

  • Ryan

    I’ve been to Europe many times and I hate roundabouts, both as a driver and as a pedestrian. The bigger ones are truly terrifying while the small ones are less effective at slowing traffic than stop signs. I think that four-way stop signs are much better because they force you to actually stop and pay attention to intersection traffic. In Europe, one of the biggest reasons for their success is because you don’t have to stop and go or slow down as in the US w/ traffic lights and stop signs and I don’t think that particularly helps our cause.

  • JohnB


    The big ones are terrifying only when you are not used to them. Once you develop a set of techniques for them, they are highly effective at getting through busy intersections. But getting to that point takes a lot of nerve for the average American encountering them for the first time.

    As for your second point i.e. that they “are less effective at slowing traffic”, I’d respond as I did before – that’s because that isn’t the purpose of them. Their purpose is to speed up traffic!

    If you actually want to slow down or stop traffic, then you’d use lights or signs which of course they often have there too. But a well-designed roundabout will process more traffic, more efficiently and more equitably than a system that forces everyone to stop whether they need to or not.

  • patrick

    JohnB, I think we must be arguing semantics, I agree wholeheartedly with that post you just made!


Proving to Caltrans That El Camino Real Can Be a Safer Street

Despite pockets of new development, El Camino Real remains a dangerous, car-oriented urban highway along most of the San Francisco Peninsula. If it can ever transform into a great street, it will have to become safer for walking and biking. And while enhancing walkability is a key goal of the Grand Boulevard Initiative — the long-term planning […]

San Mateo Adopts Vision Zero and LOS Reform With Sustainable Streets Plan

With the adoption of its Sustainable Streets Plan on Tuesday, San Mateo’s City Council has embraced several major transportation policy reforms, including Vision Zero and reforming car-oriented street design and development practices. The Sustainable Streets Plan lays out detailed design guidelines for complete streets, envisioning “a transportation system that is sustainable, safe, and healthy and supports […]

Belmont’s Ralston Corridor Study Ignores Need for Safe, Direct Bicycling

The Belmont City Council is gearing up to decide on a list of infrastructure investments intended to improve safety and reduce traffic congestion on Ralston Avenue. At a community meeting last month, representatives from consulting firms W-Trans and Alta Planning presented their Ralston Avenue Corridor Study, intended “to improve the multi-modal function” of the busy arterial street. Ralston Avenue […]

Modesto: A Model Bike-Friendly City?

The California central valley town of Modesto is not usually high on anyone’s list of cities embracing cutting-edge pedestrian and bicycle infrastructure. But that may change soon. Modesto is one of the oldest cities in California, but like other central valley towns, it grew to accommodate car travel, with wide, fast streets, plenty of parking, […]

Among Walkable Regions, San Francisco One of Most Dangerous

A pedestrian crosses Van Ness Avenue, wary of a driver who seemingly intends to switch into the left lane. Photo: Bryan Goebel Just how dangerous is San Francisco for pedestrians? A new report on pedestrian safety in the 52 largest U.S. metro areas ranked San Francisco-Oakland-Fremont 13th safest for walkers, based on an index that […]