A census of the City of San Francisco's published SFMTA parking-citation records — 23.9 million citations since 2008 — with a detailed look at the most recent twelve months: who gets cited, for what, and how heavily citation volume concentrates among repeat plates.
Trailing-year figures cover July 19, 2025 – July 19, 2026, pulled July 19, 2026 from DataSF dataset ab4h-6ztd.
San Francisco operates one of the largest municipal parking-enforcement programs in the United States. In the twelve months ending July 19, 2026, the city issued 1,500,305 parking citations carrying $166,393,564 in fines, across 531,743 distinct license plates — roughly 2.8 citations per cited plate, at an average fine of about $111.
Two findings stand out. First, volume is dominated by routine, schedule-driven violations: street cleaning alone accounts for 38% of citations, and street cleaning plus the two meter codes account for over half. Second, citations are heavily concentrated among repeat plates: 252,418 plates were cited two or more times, receiving 77.8% of all citations and 78.2% of all fined dollars. Fewer than 3,600 plates — 0.7% of all plates cited — received 1 in every 10 citations the city issued.
That pattern is characteristic of a notification and habit problem, not an enforcement gap: the same vehicles keep incurring the same predictable, avoidable violations. Section 5 discusses what that implies for compliance programs.
All figures are computed directly from the City of San Francisco's public SFMTA Parking Citations dataset, published on DataSF (data.sfgov.org) as Socrata resource ab4h-6ztd. At pull time the dataset contained 23,930,556 rows, with the earliest citation dated January 1, 2008 and rows added through July 19, 2026 — an average of roughly 1.3 million citations per year over 18½ years. Each row is one citation, with fields including citation number, issue date/time, violation code and description, fine amount, location, and the vehicle's plate and registration state as published by the city.
Aggregates were computed server-side via the Socrata Query Language (SoQL) API on
July 19, 2026 at 17:09 UTC — no bulk download, no sampling. The primary analysis window is
the trailing year: citation_issued_datetime >= '2025-07-19' AND < '2026-07-19'.
Within that window we ran:
count(*), sum(fine_amount), and
count(distinct vehicle_plate) over the full window;violation_desc, by calendar
month (date_trunc_ym), by vehicle_plate_state, and by
citation_location;group by vehicle_plate having count(*) >= 2, excluding null/blank plate
values). This returned 252,418 plate-level rows, which were then banded by
citation frequency (2–4, 5–9, 10–24, 25+).Unlike some cities' datasets, San Francisco's parking-citation dataset contains no payment-status, penalty, or balance field. Every dollar figure in this report is therefore the fine amount at issuance — what was cited, not what remains unpaid, what was collected, or what was reduced on appeal. This report measures the scale and structure of citation issuance; it cannot and does not estimate outstanding debt. (In cities that do publish balances — e.g., Baltimore and New York — the equivalent analysis can be run on open amounts directly.)
Roughly 54,000 citations in the window carry no usable plate value (consistent with the ~51,500 “NO PLATES” citations issued); these are included in citation and dollar totals but cannot appear in plate-level analyses. Violation types are published as the city's short codes (e.g., “STR CLEAN”); plain-language glosses in this report are ours. A small number of plate values are fleet or placeholder strings as published. The dataset updates continuously; figures reflect its state on July 19, 2026. Prior-period figures reused from our earlier analyses are labeled with their own computation dates.
Monthly figures are full calendar months, July 2025 – June 2026 · pulled July 19, 2026.
Citation volume is running well above its recent baseline. Calendar-year 2024 saw 1,266,807 citations and $130.5M in fines across 510,971 plates (computed July 2026 from the same dataset). The trailing year's 1,500,305 citations represent an 18% increase over the 2024 pace; the twelve full calendar months July 2025 – June 2026 total 1,550,374. October 2025 was the peak month — 201,554 citations and $22.4M in fines — with a softer early 2026 (February: 85,765).
Top 10 violation types, trailing year July 19, 2025 – July 19, 2026 · pulled July 19, 2026.
One violation type towers over everything else. Street cleaning produced 573,113 citations and $60,165,339 in fines in the trailing year — 38% of all citations and 36% of all fined dollars, more than the next six violation types combined. Add the two meter codes (238,679 citations, $25.6M) and residential overtime (119,168, $12.9M), and 62% of San Francisco's citation volume comes from three routine, time-and-schedule violations.
These are not dangerous or contested behaviors — they are calendar failures. A street-sweeping schedule the driver forgot, a meter that expired mid-errand, a two-hour zone outstayed. They are also the violations most amenable to prevention: the schedule is known in advance, the citation posts to the public dataset quickly, and the same plates incur them again and again (Section 4). Every one of these citations represents a driver who would rather have moved the car.
Top recorded citation locations, trailing year (as pulled July 19, 2026): 50 Drumm St — 8,791 citations / $2,163,370; 555 Terry A Francois Blvd — 2,093; 85 Mission St — 1,980; 501 Terry A Francois Blvd — 1,334; 3252 Pierce Street Garage — 1,320.
Volume concentrates downtown and along the Embarcadero/Mission Bay corridors — dense neighborhoods where residents park on the street and sweeping schedules rotate block by block.
Plate-frequency analysis, trailing year July 19, 2025 – July 19, 2026 · pulled July 19, 2026. Excludes citations without a usable plate value.
Grouping all 1.5M citations by plate reveals the report's central finding: San Francisco's citation volume is not spread thinly across the driving public — it is concentrated on a repeat population. Of 531,743 distinct plates cited, 252,418 (47%) were cited two or more times. Those repeat plates received 1,166,818 citations — 77.8% of the citywide total — carrying $130,042,156 (78.2%) of all fined dollars.
| Citations per plate | Plates | Citations | Share of all citations | Fines issued | Avg / plate |
|---|---|---|---|---|---|
| 2 – 4 | 185,215 | 470,621 | 31.4% | $52,694,108 | $284 |
| 5 – 9 | 45,946 | 292,462 | 19.5% | $32,512,993 | $708 |
| 10 – 24 | 17,664 | 247,257 | 16.5% | $27,401,658 | $1,551 |
| 25 or more | 3,593 | 156,478 | 10.4% | $17,433,397 | $4,852 |
| All repeat plates (2+) | 252,418 | 1,166,818 | 77.8% | $130,042,156 | $515 |
The remaining ~22% of citations comprise roughly 279,000 plates cited exactly once plus ~54,000 citations with no usable plate value. “Avg / plate” is fines issued per plate over the year. As of July 19, 2026.
67,203 plates — 12.6% of all plates cited — received five or more citations and account for 46.4% of all citations and $77,348,048 in fines. Just 21,257 plates (4.0%) were cited ten or more times, generating 403,735 citations — 27% of the citywide total. And at the extreme, 3,593 plates (0.7%) received 25+ citations each — an average of 43.5 citations and $4,852 in fines per plate in a single year.
The most-cited plate in the trailing year (masked: 7T•••93) received 314 citations totaling $34,409 — roughly six citations every week. The next most-cited (6B•••66) received 294 citations ($31,710). Several plates in this top tier show patterns consistent with commercial or fleet registrations — vehicles for which citations have effectively become an unmanaged operating cost.
The trailing year also includes 73,794 citations (~$8.0M in fines) to 28,676 plates registered outside California — led by Washington (8,131 citations), Texas (6,731), Arizona (6,398), Oregon (6,029), and Nevada (5,741). Visiting drivers are the least likely to see a mailed notice in time — a reminder that citation awareness is a cross-jurisdiction problem. As of July 19, 2026.
Read together, the two structural facts of this dataset — routine violations dominate and repeat plates dominate — point to the same conclusion: a large share of San Francisco's parking-citation volume is a notification and behavior problem, not an enforcement problem. Enforcement is clearly working: 1.5 million citations a year, posted to public data within days. What the numbers show is that the feedback loop back to the driver is not changing behavior for a quarter-million repeat plates. A driver cited 25+ times for street cleaning has not decided fines are acceptable — more often, notices arrive late, pile up unread, or go to an old address, and the pattern simply continues.
That framing matters because the two problems have different solutions. An enforcement problem calls for more officers and higher fines. A notification problem calls for reaching the right 67,000 plates, promptly, through a channel they actually see — which is dramatically cheaper, and which the concentration makes tractable: timely alerts to the 12.6% of plates with five or more citations would touch nearly half of all citation volume. The likely results are fewer repeat violations, faster payment of the citations that do occur, fewer escalations to late penalties and holds, and less frustration on both sides of the windshield.
TheTrafficApp operates a consent-based monitoring service that covers San Francisco: drivers who register a plate receive an alert when a new SFMTA citation appears in the city's public dataset, with a link to the city's official payment channels. We do not process payments or alter the official record; the service adds a timely notification layer on top of the systems the city already runs.
TheTrafficApp is a Maryland-based traffic-violation monitoring service founded by Keya Gemechu. The service monitors public citation and violation sources across multiple U.S. jurisdictions and notifies registered users — with their consent, for their own plates — when new citations appear, before late penalties accrue. Payment always happens through each jurisdiction's official channels.
The Municipal Data Series publishes independent analyses of cities' open citation datasets to give agencies, researchers, and partners a clear, sourced picture of citation scale, composition, and concentration. Each report is reproducible from the cited public dataset and query approach. We welcome corrections and collaboration: partnerships@thetrafficapp.app · thetrafficapp.app.