Helmets, lane splitting, filtering and two abreast: the law in every state and province

What these four things are

Check the states on your route

Add the states or provinces you will ride through, in order — tap them on the map, or tick them off the list. You get a card per jurisdiction, a flag at every border where a rule changes, and the tables below narrow to your route.

Every U.S. state and DC

Read across a row for the four rules; each cell links the section we read. Use the arrows in the column headers to filter — every state with no helmet law, say — and the table follows the route you picked above. A verify locally mark means we could only confirm it from an official summary, not the code itself.

Canada

The case for splitting and filtering

The argument for both is a safety argument. The time saved is real and it is not the point. A motorcycle stopped at the back of a queue has a bumper in front of it, a driver behind it who may be looking at a phone, and nowhere to go. Filtering moves the bike out of that slot. Splitting keeps it from ending up there. That is the claim. Here is what supports it, and what does not.

The California study. The one large piece of research on the practice was done by the Safe Transportation Research and Education Center at UC Berkeley and is dated May 29, 2015. CHP officers filled out a supplemental form at every motorcycle collision they investigated between June 2012 and August 2013. That produced 5,969 collision-involved motorcyclists, 997 of whom — 17 percent — were lane splitting at the moment they crashed. Berkeley no longer serves the report; California’s Office of Traffic Safety does, and that is the copy linked here.

The riders who were splitting came out of their crashes in better shape. Head injury, 9 percent against 17. Torso injury, 19 against 29. Fatal injury, 1.2 percent against 3.0. They were also less likely to have been rear-ended, 2.6 percent against 4.6 — and much more likely to have rear-ended somebody else, 38 percent against 16. That second number belongs in the same paragraph as the first.

Two findings matter more than the rest, because they are the conditions attached to everything above. Injury did not rise meaningfully until the surrounding traffic was moving faster than roughly 50 mph. And the speed differential mattered more than the traffic speed: up to 15 mph faster than the cars around them, riders’ injury rates did not move. Above that, every injury type got worse as the gap widened.

What the study does not show is that splitting is safer than not splitting, and the authors say so in as many words: the data “cannot be used to compare the collision risks for lane-splitting or non-lane-splitting riders,” because every rider in it had already crashed. It describes how riders fare once a crash happens, and under what conditions. That is a narrower claim than the one usually built on top of it, and it is the only one we will make from it.

What California actually tells riders. The state wrote lane splitting into the vehicle code in 2016 and the section took effect on January 1, 2017. No other state had such a section, and the next one to move waited until 2019. It defines splitting, it says the CHP may publish guidelines, and it sets no speed limit of its own. The safety tips the CHP publishes are shorter than the folklore around them: danger increases at higher speed differentials, danger increases as overall speed increases, it is typically safer to split between the far left lanes than between the other lanes, avoid splitting next to large vehicles, and riding on the shoulder is illegal and is not lane splitting.

You will often see a “never above 30 mph, never more than 10 mph faster” rule attributed to the CHP. Those numbers are not on the page the CHP publishes today, which carries no speed figures at all. If you want numbers, use the two the Berkeley report supports: traffic moving at 50 mph or less, and no more than 15 mph faster than it.

Five states since 2019. Every one of them put the speed cap in the statute rather than leaving it to guidance.

• Utah, 2019. Pass a vehicle that is stopped, at 15 mph or less, on a road with two or more lanes in the same direction and a limit of 45 mph or less, or on a multi-lane off-ramp. Never an on-ramp. Utah Code § 41-6a-704(6)
• Montana, 2021. The only one of the five with no road-type or speed-limit condition attached: a two-wheeled motorcycle may overtake a vehicle that is stopped or moving at 10 mph or less, traveling no more than 20 mph itself, wherever the lanes are wide enough to do it safely. MCA § 61-8-392
• Arizona, 2022. Stopped vehicle, 15 mph or less, at least two lanes in the same direction, limit 45 mph or less. A.R.S. § 28-903(F)
• Colorado, 2024. Stopped vehicle, 15 mph or less, plus one condition the others do not have: the vehicles in the adjacent lanes must be stopped too, and you stop passing the moment traffic starts moving. C.R.S. § 42-4-1503(3)(b)
• Minnesota, 2025. Up to 25 mph and no more than 15 mph faster than the traffic you are passing, with roundabouts, school zones, single-lane work zones and queued freeway on-ramps carved out. Minn. Stat. § 169.974, subd. 5

Three of those five legalized only the narrow version, filtering past traffic that has already stopped. Montana and Minnesota went further and allowed passing within the lane while traffic is still creeping, which is why those two sit in the splitting column of the table above beside California.

Two of the five are on probation. Utah’s filtering subsection repeals itself on July 1, 2027 unless the legislature extends it again, which it has already done once; Colorado’s repeals itself on September 1, 2027. Utah’s transportation committee took the question up again on May 20, 2026, with options running from making the law permanent to repealing it early. And in 2025 Utah went the other way on the wider practice, defining lane splitting for the first time and banning it, with a license suspension attached. Legal today is not the same as legal in 2028.

The part nobody writes into a statute. A rider stopped in traffic in Phoenix in July takes heat off the pavement, off the engine underneath, and off the cars on both sides, in gear that does not come off. Riders in hot states do not filter to save four minutes. They filter to stop cooking.

The honest counterpoint. Drivers do not like it. The same Berkeley report describes a 2014 survey of 951 California drivers and 709 motorcyclists in which 61 percent of the drivers “somewhat” or “strongly” disapproved of lane splitting — in the one state where it was legal at the time. That survey was never published on its own, which is a real limit on the number; the report is the only place it lives.

The confusion is worse than the disapproval. In the same survey, 36 percent of drivers believed splitting was illegal in California and another 12 percent were not sure. Nearly half the traffic, in the only state that allowed it, thought the rider coming up the line was breaking the law. A driver who believes that is a driver who may decide to correct you.

And the worry that legality invites abuse is not a silly one. The Berkeley authors found that “a significant number of motorcyclists lane-split in fast-moving traffic or at excessive speed differentials” — which is to say, in exactly the conditions their own data flags. Making a practice legal does not make people do it the safe way. Those are reasonable objections, held by reasonable people, and we are not going to pretend otherwise.

Where we stand. We think the evidence supports legal, rule-bound filtering and splitting in every state, written the way Utah, Montana, Arizona, Colorado and Minnesota wrote theirs, with the speed conditions in the statute where a rider and an officer can both read them. The alternative is not a safer rider. It is the same rider doing the same thing without guidance, or sitting in the one position on the road a rider cannot defend.

Fifteen over is permitted in Minnesota. Fifteen over is a bad idea in the rain, beside a truck, with a foot of tar snake down the lane line, or anywhere a driver has a reason to change lanes without looking. The states that legalized this did not promise anyone it was safe. They said where it was allowed. The rest is the rider’s.

Helmets: wear one, and the law is a separate question

Start with the part that is not in dispute. NHTSA estimates a helmet to be 37 percent effective at preventing death for a motorcycle rider and 41 percent for a passenger. The agency spells it out: “for every 100 motorcycle riders killed in crashes while not wearing helmets, 37 of them could have been saved had all 100 worn helmets.” That estimate is in the 2021 edition of its motorcycle fact sheet, published June 2023. The current edition, covering 2023 and published July 2025, drops the effectiveness line and reports use instead, so the older sheet is the one cited here.

Use tracks the law closely. The National Occupant Protection Use Survey, the federal roadside count, put DOT-compliant helmet use at 82.7 percent in 2023 in states that require helmets of every rider, against 65.9 percent everywhere else, with a national rate of 73.8 percent. Among motorcyclists killed that year, counting only the cases where helmet use was known, 51 percent were unhelmeted in states without a universal law and 10 percent were unhelmeted in states with one.

Michigan is the closest thing to an experiment. On April 13, 2012 the state stopped requiring a helmet of everyone and started letting riders 21 and over go without one if they carry medical coverage and have either held the endorsement two years or passed a safety course. That is still the law there, and it is the row you see in the table above. Researchers at the University of Michigan compared the twelve months on each side of that date and found helmet use among riders in crashes fell from 93.2 percent to 70.8 percent, head injury among the riders who reached a trauma center rose from 43.4 percent to 49.6 percent, and neurosurgical intervention rose from 3.7 percent to 6.5 percent. The share of crash-involved riders who died did not move, 3.3 percent against 3.2, and the authors report that plainly rather than burying it. The injury signal was clear. The death signal, in one year of one state, was not.

So: wear one. That is not a legal position, it is arithmetic, and it holds in the 2 states with no helmet law exactly as it holds in the 18 that require one of everybody.

Whether the state should require it is a different argument, and this page does not take a side in it. The strongest version of the case for a mandate is about density: a rider in heavy urban traffic meets more vehicles per mile and has less room to avoid any of them, so a universal law does more work in a congested state than the same law does where the riding is rural. The strongest version of the case against is the one California’s own highway patrol records on its motorcycle safety page: that riders advocating repeal “contend it is a matter of individual choice whether to wear a helmet or not, and a personal right to decide whether to take the risk.” That is a real argument about who owns a risk, not a claim that helmets do not work, and nobody making it is confused about the physics.

Wearing a helmet and being made to wear one are two questions. We have a position on the first one and no business settling the second. What we can do is tell you which answer the state you are entering has picked, which is the whole point of the table.

You will rarely need to split on a well-planned trip

Everything above matters most in the place a good road trip spends the least time. Lane splitting is a commuter’s tool. It solves a problem that exists on eight lanes of stopped freeway at five in the afternoon, and it has nothing to offer on a two-lane road through a national forest, where there is no second lane to split and no queue to filter past.

Plan the trip in RoadTripp and it builds days that start and finish in a real town rather than wherever the mileage happened to run out, at a daily distance you will still be enjoying at the end of it. A trip built that way spends its hours on the road and its evenings somewhere with a bed and a kitchen, which is already most of the way clear of the traffic this article is about.

If you want the road itself to be the point, that is what the Epic Trips catalog is for. Those routes are curated by hand, one at a time — Tail of the Dragon, the Beartooth Highway, Going-to-the-Sun Road, the Million Dollar Highway — and on roads like those there is nothing to split and nobody to filter past. The problem you are reading about does not exist there.

The rest is your own timing, and no app makes that call for you. If the day has a city in it, cross the city before the rush or after it. Leave at first light and take breakfast on the far side, or sit out an extra hour with a second cup and let the road clear. It costs nothing and it is worth more than any amount of filtering.

You will still meet traffic. A wreck closes a lane, a work zone backs up, a ball game lets out at the wrong moment. That is when the state you happen to be standing in matters, and it is why the table above exists. On most days of a well-planned trip, though, the answer to “can I split here?” is that there is nothing to split.

Canada in brief, and Mexico

Canada is simpler than the United States and stricter. All 13 provinces and territories require a helmet, and 4 of them carry one narrow exemption: Alberta, British Columbia, Manitoba and Ontario excuse a Sikh rider who wears a turban, with Alberta and Ontario setting that exemption at 18 and over, and Manitoba adding a medical exemption certificate and a lawfully authorized parade. Lane splitting is legal in 0 of the 13, and filtering in 0. Only Quebec bans riding between rows of vehicles in so many words; everywhere else the work is done by the ordinary rule that a driver stays in one marked lane and changes lanes only when it is safe, which is why those rows cite a lane-discipline section rather than a motorcycle one. Two abreast is the rule that surprises Americans going north: where most U.S. states allow it, seven Canadian jurisdictions require single file and only Saskatchewan says plainly that two motorcycles may share a lane.

Mexico has no rows in the table and will not get any until we can read its rules the way we read these. Helmet requirements are set state by state rather than federally, the official pages that publish them are uneven and often undated, and lane splitting there is less permitted than unregulated: it happens constantly, and we could not find a statute to point you at that says it is allowed. Ride as though nothing is permitted. Helmet on everywhere, and do not assume that what is legal in California is legal, or advisable, in Sonora. If we find primary sources we can stand behind, Mexico gets rows like everyone else.