Red Light Cameras in 2026: Where They Are Legal, What a Ticket Costs, and What the Evidence Shows

{
“@context”: “https://schema.org”,
“@type”: “Article”,
“headline”: “Red Light Cameras in 2026: Where They Are Legal, What a Ticket Costs, and What the Evidence Shows”,
“description”: “Which states allow red light and speed cameras, what a camera ticket means, California AB 645, and what 893 contracts show about who runs these programs.”,
“author”: {
“@type”: “Person”,
“name”: “Abbas Khan”
},
“publisher”: {
“@type”: “Organization”,
“name”: “Civic IQ”
},
“datePublished”: “2026-09-16”,
“dateModified”: “2026-09-16”,
“mainEntityOfPage”: {
“@type”: “WebPage”,
“@id”: “https://civiciq.com/blog/red-light-cameras”
}
}

{
“@context”: “https://schema.org”,
“@type”: “FAQPage”,
“mainEntity”: [
{
“@type”: “Question”,
“name”: “How much is a red light camera ticket?”,
“acceptedAnswer”: {
“@type”: “Answer”,
“text”: “It depends on the state. California set a graduated civil schedule in Vehicle Code 21455.9, effective January 1, 2026. It runs $100 for a first offense, then $200, $350 and $500 within three years. Under the AB 645 speed pilot, fines start at $50.”
}
},
{
“@type”: “Question”,
“name”: “Do red light camera tickets put points on your license?”,
“acceptedAnswer”: {
“@type”: “Answer”,
“text”: “Usually not. In most states it is a civil penalty against the registered owner. It never reaches your driving record, so it does not normally affect insurance. A few states do treat some camera citations as moving violations. Check your own.”
}
},
{
“@type”: “Question”,
“name”: “Can you ignore a red light camera ticket?”,
“acceptedAnswer”: {
“@type”: “Answer”,
“text”: “That is risky and varies by state. In some places an unpaid citation goes to collections or blocks a registration renewal. Read the document you received: a notice of violation and a formal summons are different things with different results.”
}
},
{
“@type”: “Question”,
“name”: “How do you fight a red light camera ticket?”,
“acceptedAnswer”: {
“@type”: “Answer”,
“text”: “Request the photo and video evidence first. The timestamp shows how long the light had been red, which is the central fact. Common winning challenges are simple. A misread plate. A vehicle already sold. A legal right on red. Or a yellow interval below the state minimum.”
}
},
{
“@type”: “Question”,
“name”: “Which states ban red light cameras?”,
“acceptedAnswer”: {
“@type”: “Answer”,
“text”: “Idaho, Maine, Mississippi, Montana, New Hampshire, South Carolina, South Dakota, Texas and West Virginia have express prohibitions. Texas ended its programs in 2019 but let pre-ban contracts run out, so Leon Valley remained on the IIHS list.”
}
},
{
“@type”: “Question”,
“name”: “Are speed cameras legal in California?”,
“acceptedAnswer”: {
“@type”: “Answer”,
“text”: “Yes, but only in a narrow pilot. AB 645 authorizes them in six cities: Los Angeles, San Jose, Oakland, Glendale, Long Beach and San Francisco. A separate law covers Malibu’s stretch of Pacific Coast Highway. The pilot is repealed on January 1, 2032.”
}
},
{
“@type”: “Question”,
“name”: “Who operates red light cameras?”,
“acceptedAnswer”: {
“@type”: “Answer”,
“text”: “Private vendors, under multi-year deals with the agency. Civic IQ reviewed 893 priced camera contracts. Verra Mobility appears most often, at 140. Redflex follows at 82 and Sensys Gatso at 60.”
}
},
{
“@type”: “Question”,
“name”: “Do red light cameras actually reduce crashes?”,
“acceptedAnswer”: {
“@type”: “Answer”,
“text”: “Yes, with a tradeoff. The FHWA seven-city study found right-angle crashes down about 25% and rear-end crashes up about 15%. The trade is favorable because a side impact is far more likely to kill someone than a rear-end hit.”
}
}
]
}

Last updated: September 16, 2026

If a ticket arrived in the mail, the section on what a camera citation actually is will answer most of it. If you want to know whether these systems work, the honest answer is more mixed than either side usually admits.

---

How does a traffic enforcement camera work?

A red light camera fires on a trigger. Sensors buried in the pavement, or radar, detect a vehicle entering the intersection after the signal has already turned red.

It does not fire on vehicles that enter on yellow. That distinction is built into the system.

The camera captures a package rather than one photo. That usually means a shot at the stop line, a second shot mid-intersection, a rear plate image and a short video clip. Each one is stamped three ways: the signal phase, the time since the light turned red, and the vehicle speed.

Speed cameras use the same idea with a different trigger. Radar or lidar measures speed against the posted limit. The system only captures above a set point. That point is usually 10 or 11 mph over. It is set high on purpose, to avoid arguments about speedometer accuracy.

A human still makes the call

The evidence package goes to a reviewer. Most places run two passes.

A vendor technician screens out the obvious ones first. That means emergency vehicles, funeral processions and plates nobody can read. A sworn officer or approved agency employee then decides whether to issue.

Most states require that second review by law. Programs that skipped it have lost in court.

One design detail surprises most people. These systems shoot the rear plate rather than the driver. That is why the ticket goes to the registered owner, not to whoever was driving.

---

Where are red light cameras and speed cameras legal?

There is no national law on camera enforcement. Every rule is state law, and the map is patchy.

Nine states ban red light cameras. New Hampshire and West Virginia went first in 2006. Mississippi, Montana and Maine followed in 2009, South Carolina in 2011, South Dakota in 2014, Texas in 2019 and Idaho in 2025.

Nine states ban speed cameras: Idaho, Maine, Mississippi, New Hampshire, New Jersey, South Carolina, Texas, West Virginia and Wisconsin.

The gaps are worth noticing. Montana and South Dakota ban red light cameras but have no speed camera law at all. New Jersey and Wisconsin are the reverse. New Jersey banned photo radar back in 1992, and its red light pilot just expired in December 2014.

Permitted rarely means statewide

Only five places allow speed cameras generally: Arizona, Colorado, DC, Iowa and Oregon.

Everywhere else, lawmakers approved them inside narrow boxes. School zones only in Florida, Georgia and North Carolina. Work zones only in Indiana, Kentucky, Michigan and Vermont. Rhode Island allows school zones on school-year weekdays, 7am to 6pm. California, Pennsylvania, Minnesota, New York and Maryland name specific cities.

Many carry expiry dates. Minnesota ends in 2029. Vermont sunsets in October 2026. Pennsylvania red light authority expires July 15, 2027 unless the legislature extends it.

A state can permit cameras and still make them unworkable

Ohio allows both camera types in municipal corporations. It then holds back local government fund money from cities that use them. The Ohio Supreme Court upheld that in May 2022.

Opponents elsewhere took the hint. A proposed Ohio bill would add a $100,000 yearly fee per device, plus an 8% tax on vendor revenue. A ban is no longer the only tool.

Texas shows the same idea from the other side. HB 1631 ended red light cameras in 2019. Cities already under contract were allowed to run those out. Leon Valley is the last Texas city on the IIHS list.

---

What is a red light camera ticket, legally?

In most states it is a civil penalty rather than a moving violation. That distinction carries four real results.

• No points on your license. No officer identified the driver, so most states treat it as an owner-liability matter.
• Usually no insurance impact, because it never reaches your driving record.
• The registered owner is presumed responsible. Most states let you rebut that with an affidavit naming the driver, or by showing the vehicle was stolen or sold.
• Fines are typically lower than an equivalent officer-issued ticket, precisely because the penalty is civil.

California wrote that model into Vehicle Code 21455.9, effective January 1, 2026. Red light violations became civil offenses, with no points and no suspension. The schedule runs $100, then $200, $350 and $500 within three years.

What varies dangerously by state

Whether ignoring it has teeth. In some places an unpaid citation goes to collections, blocks a registration renewal, or reaches a court. In others there is almost no enforcement mechanism.

Some programs send a notice of violation that looks like a citation but carries no legal force until formally served. Read what arrived carefully. A notice and a summons are different documents.

Whether points apply after all. A minority of states do assess them. Arizona and Oregon have in the past treated some camera citations as moving violations requiring personal service.

The contest window. Most programs give 30 to 45 days and let you request the photo and video first. Requesting it is almost always worth doing, because the timestamp shows how long the light had been red.

The most common successful challenges are mundane. A misread plate. A vehicle already sold. A legal right on red where the local ordinance permits it. Or a yellow interval shorter than the state minimum.

That last one matters more than it sounds. Yellow timing is the single biggest driver of red-light-running rates. Programs that shortened yellows to lift citation counts caused most of this industry’s scandals.

---

California speed cameras: what AB 645 actually does

California banned speed cameras for decades. AB 645 changed that in 2023. It is a tight pilot, set out in Vehicle Code sections 22425 to 22431. Other states are already borrowing from it.

The two provisions that matter most

The anti-bounty clause. AB 645 bars paying the camera vendor per citation, or as a share of revenue. That setup has driven most of the public backlash against these programs. Washington State has a similar rule.

There is a gap in it. The statute still permits a separate notice-processing vendor to be paid per notice processed.

The equity mandate. Cities must publish an impact report. It breaks violations down by ZIP code and by income status, and is built with local racial justice and equity groups. If cameras cluster in low-income neighborhoods, the city has to justify it.

Revenue above program cost must go to traffic calming within three years. If not, it goes back to the state.

Where each city stands as of September 2026

One buying detail is worth flagging. Los Angeles awarded its contract by piggybacking on Oakland’s deal. It did not run its own competitive bid.

Two other laws sit alongside the pilot. SB 1297 approved cameras on Malibu’s 21-mile stretch of Pacific Coast Highway. AB 645 could not cover it, because PCH is a state route. AB 289, from October 2025, lets Caltrans put up to 35 systems in highway work zones.

The early California results

San Francisco published a one-year review in May 2026. It reported a 79% drop in vehicles going 10 mph or more over the limit at camera sites. The share of drivers over by 10 or more fell from 11% to 2%. Sixty-five percent of drivers who got a warning or ticket never did it again.

Oakland’s warning-phase data says more about who speeds. In roughly five weeks the city issued 140,445 warnings. 35.6% of plates were repeats, and those repeat vehicles generated 66.1% of all warnings.

Speeding is clustered in a minority of drivers. That is the strongest argument for camera enforcement. It is also the strongest argument that a flat fine is the wrong tool.

One caution applies to all of these figures, including the 79%. They measure speeding behavior at treated sites, before and after, with no control group. They are not causal estimates of crash reduction. Do not compare them with the crash studies below.

---

Do traffic enforcement cameras actually work?

The research supports a narrower claim than advocates make. It also supports a far stronger one than opponents concede.

Red light cameras: yes, with a real tradeoff

The most-cited study is Hu and Cicchino (2017), covering 117 large US cities. Turning cameras on cut the fatal red-light-running crash rate 21%. It cut all fatal crashes at signalized junctions by 14%.

The sharper finding came from cities that switched cameras off. Fourteen cities shut programs down between 2010 and 2014. In those, the fatal red-light-running crash rate ran 30% higher than expected.

That 30% is widely quoted. It deserves a caveat that rarely travels with it. The figure comes from a sample of just 14 cities. Across all 19 shutdown cities the rise was smaller, and it did not reach statistical significance.

A 2020 Campbell Collaboration review covered 38 studies. It found injury crashes down 20%, and right-angle injury crashes down 29%.

The tradeoff is real and well replicated. The FHWA seven-city study found red light cameras cut right-angle crashes 25% while increasing rear-end crashes 15%.

The case for cameras does not depend on that rise not happening. It rests on severity. A side impact is far more likely to kill someone than a rear-end hit. The same study found a net benefit above $18.5 million.

Anyone citing the crash drop without the rear-end increase is selling something. So is anyone citing the rear-end increase without the severity difference.

Speed cameras: yes, but the headline numbers are inflated

The Cochrane review of 35 studies found drops across every measure. Average speed fell 1 to 15%. Injury crashes fell 8 to 50%. Fatal and serious injury crashes fell 11 to 44%.

The authors would not compute a single pooled figure. They cited wide variation and weak methods. The review is also sixteen years old now, and has not been updated.

The most useful paper for anyone evaluating camera claims is Graham and colleagues (2019) in PLOS ONE. It looked at 771 fixed camera sites across eight English districts. The method was a doubly-robust causal estimator.

Roughly half to two-thirds of the naive headline effect is selection bias. Cameras get sited where crash counts just spiked, and those counts would have partly fallen anyway. That is regression to the mean.

A real causal effect near 15% is still large, and statistically solid. But treat any study claiming 30% or more with care. Without a control group or an empirical Bayes adjustment, it is overstating its case.

Recent US work matches the adjusted figure. A 2025 study covered more than 1,800 New York City school zone cameras. It found a 14% crash drop and a 75% drop in speeding. Cameras took about six months to reach full effect, so behavior change is gradual.

FHWA lists speed safety cameras as a Proven Safety Countermeasure, though it stops short of endorsing a single pooled effect size.

The equity question, and why two literatures disagree

Two credible bodies of research look like they contradict each other. They do not. They answer different questions.

The burden is spread unevenly. A D.C. Policy Center analysis looked at violations per person. Drivers in Black-segregated census tracts got about twice the average. White-segregated tracts got about an eighth. That is a 17-fold spread, even though crash rates per person were similar. ProPublica found Chicago households in Black and Latino ZIP codes ticketed at roughly twice the rate of white ZIP codes.

The citation decision itself is less biased than an officer’s. A 2024 PNAS study made a crucial methodological move. It used the estimated racial mix of actual road users as the denominator rather than residential population.

On that basis, camera citations tracked who was actually driving. In Chicago neighborhoods where about half of drivers were Black, Black drivers made up 70% of police stops.

Both findings are true. Where cameras go, and flat fines, are regressive. The moment-of-citation decision is not. That split is what AB 645’s ZIP-code reporting and income-based discounts are built to address. No California city has yet published data on how many drivers use them.

---

Who actually operates these programs?

Almost no city runs its own cameras. Private vendors operate them on multi-year deals. The market is clustered.

The public numbers say Verra Mobility leads. It reported $979.1 million in FY2025 revenue, with Government Solutions at about $505 million. Its 2021 purchase of Redflex merged the two largest US operators into one.

The contract records say the same thing, and they let you measure it. Across 893 priced camera contracts surfaced by Civic IQ from agency records, here is how often each operator appears.

Redflex still appears 82 times despite the acquisition, because legacy contracts keep running under the original name. Anyone counting market share by brand will undercount Verra.

One deal shows the scale. In 2026 New York City DOT finalized a five-year, $998 million contract with Verra Mobility. That is 34% larger than the last one. It expands red light cameras from 150 to 600 intersections, and adds speed, bus lane and weight enforcement.

How the vendor gets paid is the real signal

Two payment models dominate. The choice tells you more about a program’s integrity than almost anything else.

A fixed cost per camera per month gives the vendor no stake in citation volume. Los Angeles is the best-documented current benchmark at $4,450 per system per month.

Revenue share or per-citation payment does the opposite. It gives the supplier a direct stake in how many tickets get written. In the past it also shaped the engineering choices behind that number.

The records show both models still in use.

Fixed fee outnumbers revenue share five to one. That is the healthier picture. The revenue-share deals that remain are open about it. One budget document lists vendor fees for speed cameras as a 39% revenue share, estimated at $750,000.

California and Washington have banned per-citation payment outright. The Fines and Fees Justice Center calls it a core contract red flag. So are vendor-picked camera sites and missing human review.

---

How do agencies buy camera programs?

Mostly by renewing. Renewal or extension language appears in 120 of the 893 priced contracts. That beats every other route.

Competitive bids are more common here than in most categories we track, at 65. Camera contracts are large and public. They draw a formal process.

The awards are large. Set aside statewide deals above $50 million and 403 city awards remain. The median is $300,102, and a quarter run above $1.65 million.

Buying has also sped up sharply. The records show 19 priced camera contracts in 2019, and 302 in 2025.

---

Where is this heading?

Three signals matter, and they point in different directions.

Speed cameras are expanding, narrowly. Every new state approval since 2023 has been for speed cameras. Michigan and New York came in 2024, North Carolina and Kentucky in 2025. Maryland, Washington, Colorado, Virginia and California expanded theirs. Nearly all of it is limited to school zones and work zones, with sunsets attached. Louisiana actually narrowed its law in 2025.

Red light cameras are contracting. No state has newly approved them in years. The movement has run the other way. Texas banned them in 2019 and Idaho in 2025. Pennsylvania authority sunsets in July 2027.

The opposition changed tactics. Instead of bans, opponents now use rules that make programs unworkable. Georgia’s HB 651 took effect in July 2026. It overhauled school zone camera rules and requires a local voter referendum for programs starting July 1, 2027.

That converts one legislative fight into hundreds of local ones, which is the point.

The churn is enormous relative to the change. MultiState counted roughly 300 camera enforcement bills introduced in 2025 and just over 30 enacted, a 10% conversion rate.

Some programs are also ending on their own terms. Avondale, Arizona voted in November 2025 to shut its speed cameras down, ending in June 2026. Politics was not the reason. The city’s own review found average speeds among speeding drivers were the same before, during and after.

---

Frequently asked questions about red light cameras

How much is a red light camera ticket?

It depends on the state. California set a graduated civil schedule in Vehicle Code 21455.9, effective January 1, 2026: $100 for a first offense, then $200, $350 and $500 for later offenses within three years. Under the AB 645 speed camera pilot, fines start at $50.

Do red light camera tickets put points on your license?

Usually not. In most states a camera citation is a civil penalty against the registered owner, so it does not reach your driving record and does not normally affect insurance. A minority of states treat some camera citations as moving violations, so check your own state.

Can you ignore a red light camera ticket?

That is risky and varies by state. In some places an unpaid citation goes to collections or blocks a registration renewal. Read the document you received: a notice of violation and a formal summons are different things.

How do you fight a red light camera ticket?

Request the photo and video evidence first. The timestamp shows how long the light had been red, which is the central fact. Common successful challenges are a misread plate, a vehicle already sold, a legal right on red, or a yellow interval below the state minimum.

Which states ban red light cameras?

Idaho, Maine, Mississippi, Montana, New Hampshire, South Carolina, South Dakota, Texas and West Virginia have express prohibitions. Texas ended its programs in 2019 but let pre-ban contracts run out, so Leon Valley remained on the IIHS list.

Are speed cameras legal in California?

Yes, but only in a narrow pilot. AB 645 authorizes them in six cities: Los Angeles, San Jose, Oakland, Glendale, Long Beach and San Francisco. A separate law covers Malibu’s stretch of Pacific Coast Highway. The pilot is repealed on January 1, 2032.

Who operates red light cameras?

Private vendors, under multi-year deals with the agency. Across 893 priced camera contracts reviewed by Civic IQ, Verra Mobility appears most often at 140, ahead of Redflex at 82 and Sensys Gatso at 60.

Do red light cameras actually reduce crashes?

Yes, with a tradeoff. The FHWA seven-city study found right-angle crashes down about 25% and rear-end crashes up about 15%. The trade is favorable because a side impact is far more likely to kill someone than a rear-end hit.

---

The short version

• Speed cameras now outnumber red light cameras, 375 communities to 357. Both counts are September 2026.
• Nine states ban red light cameras. Nine ban speed cameras. Only five places allow speed cameras statewide.
• A camera ticket is usually a civil penalty against the registered owner, with no points. Check whether you got a notice or a summons. Request the photo and video before you decide anything.
• California’s AB 645 pilot is live in San Francisco and Oakland. It is pending in Los Angeles and Long Beach, and unconfirmed in Glendale and San Jose. It sunsets on January 1, 2032.
• Red light cameras cut right-angle crashes about 25%. They raise rear-end crashes about 15%. That trade is worth it because of severity.
• Speed cameras cut crashes about 15% once you correct for regression to the mean.
• Verra Mobility appears in 140 of 893 priced camera contracts, more than anyone else. How the vendor gets paid matters more than which vendor it is.

Three related reads. Our analysis of what cities spend on traffic safety analytics. Our read on Flock Safety’s plate-reader contracts. And our guide to selling public safety tech to local government.

Program counts from IIHS as of September 2026. California provisions from Vehicle Code sections 22425 to 22431 and 21455.9. Contract figures come from 893 priced camera records surfaced by Civic IQ from agency agendas, purchase orders and award notices. Effect sizes are cited from the underlying peer-reviewed studies rather than secondary summaries.