It’s been almost a year since San Francisco permanently closed the Upper Great Highway to cars and transformed the two-mile stretch of road into a park.
Soon after last March’s closure, the quest was on to measure its effect on Sunset District traffic. Until now, the most recent report came only four months later — a period of time that the city’s streets agency and independent traffic researchers say isn’t long enough for drivers, pedestrians, and cyclists to adjust to big changes.
As the anniversary approaches, foes of the closure are gathering signatures for a November ballot measure. They’ll almost certainly hit their goal and force another citywide vote on the issue two years after 55 percent of SF voters approved the closure.
There have already been claims about traffic safety and speeds, and there will be more. The Sunset’s supervisor Alan Wong, who wants to reopen the road to cars, tried this winter for his own ballot measure and failed.
He said crashes had increased in his district. A rival for his seat recently said reopening the Great Highway would improve safety and that “traffic speeds have gone up” since the closure, though he didn’t provide evidence.
The Frisc has done the first independent analysis of both serious crashes (injuries and fatalities) and congestion, with enough time elapsed since the closure for more established traffic patterns to emerge.
There was a brief post-closure spike in crashes that bolstered Wong’s claim, but the claim falls apart with a longer view. And the broad declaration that speeds are up in the Sunset isn’t true. On many roads, traffic is slower during key driving hours, often by incremental amounts.
There’s no guarantee these patterns will sustain themselves. Other factors, such as a surge in commuters, public transit service changes, more autonomous vehicles, and nearby road work could influence future behavior.
But it’s important to fact-check and add context to claims that continue to circulate as San Francisco heads toward yet another Great Highway vote.
Sunset crashes
In January, Sup. Wong said at a town hall meeting that he would push for a measure to reopen the Upper Great Highway during weekdays. He cited injury collision statistics comparing the first five months after the closure — April through September 2025 — with the same period in 2024. He said traffic injuries in the year-over-year comparison had jumped 81 percent.
Crashes did increase in that period. (We found a 71 percent jump.) But the rise was driven by a single month: 28 crashes in May 2025. It was the highest total since 31 in January 2019, before there were slow streets or permanent street closures.
But for the rest of 2025 — June through December — the monthly average was less than 10, lower even than the same period in 2020 at the depth of the pandemic. January 2026 figures were just released: there were five injury crashes in District 4.
Other than one outlier month, there’s no evidence of increased street mayhem around the closed Great Highway. In fact, it’s the opposite.
At the time of his presentation, Wong had access to October and November crash data but didn’t include them.
The SF Municipal Transportation Agency typically starts to evaluate the effect of street changes three to six months after implementation. The practice makes sense, says San Jose State University transportation researcher Marcel Moran: “Changes to underlying traffic and safety patterns can take a long time to stabilize into a new equilibrium.”
For a fuller picture of Sunset District traffic safety, The Frisc analyzed city crash data from 2019, the last full year before the pandemic, through 2025. For a simpler visual, we are only displaying averages from 2019, 2024 — the final full year before the Great Highway’s permanent closure — and 2025. (The years 2021, 2022, and 2023 saw increases in collisions due to higher rates of unsafe driving post-pandemic, though 2023 marked the start of a gradual drop in crashes that has continued.)
We supplemented the District 4 data with collisions on the northbound side of 19th Avenue, which is technically in District 7 but is a major part of the west side’s traffic pattern.
Included in these tallies are fatal collisions. There were two fatalities each in D4 in 2019 and 2025. In 2024, when citywide traffic deaths hit a 10-year high, there were four fatalities in the district.
Sunset congestion
Could the drop in crashes since May 2025 stem from slower street conditions? Are the thousands of drivers who used the Upper Great Highway every day now clogging other arteries? Slower speeds might be frustrating, but they also give drivers more time to avoid collisions.
Based on our analysis, some streets are now slower, and some are not. Albert Chow, who helped spur the recall of Sup. Joel Engardio, is one of four candidates trying to oust Sup. Wong in a special June election. At last week’s candidate forum, Chow responded to a question about improving traffic safety with a blunt response: “Reopen the Great Highway.” He also said that with thousands of drivers now diverted to other streets, “traffic speeds have gone up” across District 4.
In fact, most street segments we analyzed saw a slowdown in the morning and evening commute times. But in many cases, the change has been slight — less than 5 percent. For a person driving 25 mph, that’s about 1 mph slower. At 30 mph, it’s 1.5 mph slower.
A few chokepoints stick out. During the evening commute hours, drivers cutting through Golden Gate Park to and from the Richmond District have clogged Chain of Lakes Drive between MLK Drive and JFK Drive. Traffic is 18 percent slower northbound and 16 percent southbound, which hat translates into 2 to 3 mph slower.
SFMTA is aware of the Chain of Lakes problem, which has been a bottleneck for years. The agency said last summer that the road had 2,000 more vehicles per day than pre-closure. SFMTA has adjusted traffic signals and flow to encourage drivers to go around the park via the Great Highway between Fulton Street and Lincoln Way instead of cutting through it.
Another notable slower road is the farthest west segment of Vicente Street during evening hours. Both directions saw traffic moving 1.9 to 2.3 mph slower after the Great Highway closed. (Vicente is not shown on our maps.)
For our congestion analysis, The Frisc used San Francisco Transportation Authority (SFCTA) data for typical weekday traffic speeds — Tuesday, Wednesday, Thursday — in the morning and evening. We looked at District 4 streets in the 12-month period before the Great Highway closed (Mar. 2024 to Feb. 2025) and the 11-month period after (Mar. 2025 to Jan. 2026). As with the crash data, we also included weekday speeds for northbound 19th Ave. in District 7.
A handful of streets have gotten less congested. At the top of this list is southbound Great Highway from Balboa Street to Lincoln. During peak evening hours, traffic moved 3.2 mph faster than pre-closure.
SFCTA acting co-deputy director of technology, data, and analysis Drew Cooper notes that traffic also has an ebb and flow generally independent of particular street changes. It tends to move faster in the winter — “bad weather affects people’s willingness to drive” — slower in the spring, faster in summer when people are on vacation, and slower in the fall with people back to work and school.
Cooper also warned about other factors, like road repairs. Sewer work and repaving along 44 blocks of Sunset in 2024 had an impact. This spring, 19th Avenue between Holloway and Lincoln will be repaved, impacting travel times, particularly in the morning.
Assuming the signature gathering succeeds, SF will vote in November — the fifth major vote in four years on the Great Highway’s fate.
In 2022, the Board of Supervisors voted to create a compromise pilot, open to cars only on weekdays, for three years. The same year, SF voters rejected an attempt to reopen the road and to halt permanent closure of its crumbling southern extension. Then came 2024’s Prop. K, which in turn led to the recall of its main champion, Sup. Joel Engardio. Roughly two-thirds of Sunset District voters marked “No” on Prop K, and about the same percentage ousted Engardio.
With nearly all the special election candidates lined up against the road’s closure, the next real referendum will be the November ballot itself. Who knows if it will be the last one.