A woman stands on the pavement beside her bicycle on a quiet residential street in warm morning light, helmet in one hand.

Struck while walking or cycling: the order the questions arrive in

One ordinary claim by a person who was on foot or on a bicycle, walked through end to end — which policy answers when the injured person owns no car, what happens when the driver is uninsured or never identified, and where the fault question finally arrives.

Updated September 22, 2026 Intermediate
How a car insurance claim actually worksFull transcriptSubtitles are on by default; the player's CC button turns them off.

A crash involving someone who was walking or cycling opens a file with a hole in the middle of it. There is no vehicle of theirs to photograph, no policy of theirs to notify, no excess, no hire car. The injured person is nobody’s policyholder in the event. Whatever arrives arrives through a contract somebody else bought, or through a fund that exists because nobody bought one — and working out which of those it is comes before anybody has said a word about fault.

So this piece walks one such claim through in the order the questions actually arrive. The claim is a genre rather than a case: no name, no date, no street, no injury and no sum appear in it, because inventing any of those would make it vivid and make it false. What is specific is the instrument that answers each question, quoted from the text that was fetched and read.

The first question is not who was at fault

It is whether any policy reaches a person who was not in a vehicle at all. Across the European Union that answer is written into the compulsory insurance itself: the codified motor insurance directive says the insurance «shall cover personal injuries and damage to property suffered by pedestrians, cyclists and other non-motorised users of the roads who, as a consequence of an accident in which a motor vehicle is involved, are entitled to compensation in accordance with national civil law».

Then the clause does something easy to read past. It closes: “This Article shall be without prejudice either to civil liability or to the quantum of damages.” The cover is made to reach you, and nothing has been decided by its reaching you. That separation is the shape of the whole file, and it is why the early questions and the fault question run on different clocks.

A benefit that arrives before the fault question is answered

Several American no-fault statutes go a step further and put a first-party medical benefit in the hands of a struck non-occupant. Kansas is the plainest. Every motor vehicle liability policy issued in the state must “include personal injury protection benefits to the named insured, relatives residing in the same household, persons operating the insured motor vehicle, passengers in such motor vehicle and other persons struck by such motor vehicle and suffering bodily injury while not an occupant of a motor vehicle”.

Read that last clause on its own. The person the policy pays has no relationship to the insurer, agreed to nothing, and is described only by where they were standing.

North Dakota’s chapter does it twice over, from two directions. The insurer of a secured vehicle pays for “accidental bodily injury sustained by any pedestrian in this state as a result of being struck by the secured motor vehicle” — that is the striking vehicle’s policy. And the same chapter covers the owner of an insured vehicle, and the owner’s relatives, injured “while a pedestrian as the result of being struck by a motor vehicle or motorcycle” — that is your own car’s policy, answering for you while the car sits at home.

So a person walking to the shops can sit inside two first-party benefits at once without owning anything relevant to the crash. Whether a benefit of that kind exists where you were struck — whether the place runs a no-fault system at all — is in the jurisdiction notes below.

Whose policy, when more than one could answer

This is the hardest question on the file, and Minnesota’s no-fault chapter is one of the few instruments that answers it as an ordered list rather than leaving it to argument. Its priority section works down: for an injured person who is an insured, the benefits come from “the security under which the injured person is an insured”; for a driver or occupant of an involved vehicle who is not an insured, from “the security covering that vehicle”; and then the clause written for the person on the pavement — for “a person not otherwise covered who is not the driver or other occupant of an involved motor vehicle”, from “the security covering any involved motor vehicle”.

The same clause carries a limit that decides real cases: “An unoccupied parked vehicle is not an involved motor vehicle unless it was parked so as to cause unreasonable risk of injury.”

Note the order. Someone on foot who happens to hold a policy of their own is sent to it first, and only a person with nothing of their own is routed to the vehicles that were there. That will feel wrong to somebody who did not cause the crash and is being pointed at their own insurer. It is not a judgement about blame; it is an allocation rule about who processes the paperwork, and the chapter says so in the next subdivision: where two obligations apply, “benefits are payable only once and the reparation obligor against whom a claim is asserted shall process and pay the claim as if wholly responsible”, and that insurer then recovers a share from the others.

The argument between insurers happens, in other words, without the injured person in the room.

When the driver is uninsured, or never identified

The two cases feel the same from the kerb and are drafted separately.

In the European Union each Member State must “set up or authorise a body with the task of providing compensation, at least up to the limits of the insurance obligation” for damage caused by an unidentified vehicle or by one that was not insured. The directive then protects the route to it: the victim “may in any event apply directly to the body”, which must give “a reasoned reply”, and Member States “may not allow the body to make the payment of compensation conditional on the victim establishing in any way that the person liable is unable or refuses to pay”.

Where the two cases part company is property damage, which the directive lets a Member State limit or exclude when the vehicle was unidentified. Personal injury is not treated that way.

Spain’s consolidated text shows what the obligation looks like once implemented. The Consorcio de Compensación de Seguros is given the job of compensating personal injury «por siniestros ocurridos en España, en aquellos casos en que el vehículo a motor causante sea desconocido», and a separate paragraph of the same article covers the vehicle that was simply not insured.

The United States has no single national counterpart. Some no-fault states run an assigned claims plan instead: North Dakota’s exists so that an eligible injured person may obtain benefits through the plan where benefits “are not applicable to the injury” for a reason other than the chapter’s own exclusions, with the plan then subrogated to that person’s rights. Whether the place you were struck in requires an uninsured motorist cover on every policy, and whether it reaches a person who was not in a vehicle, is in the jurisdiction data below.

Where the fault question finally arrives

It arrives last, and where it arrives it is a question about apportionment rather than about character.

Spain’s text is unusually explicit about the starting point. The driver answers “en virtud del riesgo creado por los hechos de la circulación”, and for personal injury is released from that liability only on proving that the harm was due to «la culpa exclusiva del perjudicado o a fuerza mayor extraña a la conducción o al funcionamiento del vehículo» — the same sentence ruling out defects and mechanical failures as force majeure.

Short of that, the article provides for reduction rather than refusal: where a victim capable of civil fault only contributes to the harm, the indemnities are reduced «en atención a la culpa concurrente hasta un máximo del setenta y cinco por ciento». And then a limit on the limit, which is the part worth carrying away: for sequelae and temporary injuries, the exclusive or concurrent fault of non-driving victims under fourteen, or of victims whose impairment deprives them of the capacity for civil fault, «no suprime ni reduce la indemnización».

That tells a reader nothing about what a particular file will do with a particular crossing. It tells them what kind of question is being asked — a proportion, bounded, and in some cases barred from being asked at all. Which shared-fault rule applies where you were struck is in the jurisdiction notes below.

What a model of the impact can and cannot settle

Some of the dispute on these files is about speed and geometry, and the section’s crash simulator on a marked-crossing scenario exists to make that part legible rather than to resolve it. For a struck pedestrian it reports the impact speed and an average fatality risk computed from a published curve, plus where that speed sits against published severe-injury levels. For a cyclist it reports the impact speed and stops there — no openly readable cyclist speed–risk curve could be located, and the tool prints that absence instead of filling it. For people inside vehicles it reports an injury-risk band from the change in velocity for frontal and rear-end impacts only, and refuses to borrow the frontal curve for a side impact.

Every one of those figures is a population statistic. It is crash physics, not a prediction about one person’s body, and no number it produces describes the reader.

The same model set at night on an unlit stretch holds the sight line rather than the speed. What could be seen, from where and from what distance, is what most of these files eventually turn on — and it is the question whose evidence sits mostly on the other side.

What we cannot tell you

We could not read the United Kingdom’s untraced-driver and uninsured-driver agreements. The Motor Insurers’ Bureau operates both, and no published copy of either could be opened at a readable address on the day this was written, so nothing is said here about British conditions, time limits or deductions, and no second-hand description of them has been substituted.

We cannot tell you what any claim will produce. This section does not price claims and carries no averages, no ranges and no examples of what anybody received.

And we cannot tell you whether the vehicle’s own record will ever reach your file. A modern car may hold data about speed, braking and what its sensors registered before contact; a person standing at the kerb holds a memory. No instrument read here addresses that asymmetry.

The part that decides the shape of everything else

Which policy answers is not a detail to settle once the medical side is clearer. It decides who takes the first call, which clock starts, and whose decision has to be argued with. Every instrument quoted here answers it before it answers anything about blame.

Rules in your jurisdiction

Deadlines, fault rules and minimum coverage differ by state and country. Pick yours to see the rules that apply to this topic.

Select a jurisdiction to see its rules.

Frequently asked questions

I do not own a car and I hold no insurance of any kind. Is there anything for me to claim on?

The instruments read for this article are drafted on the assumption that there will be. The European Union's codified motor insurance directive requires the compulsory insurance to cover «personal injuries and damage to property suffered by pedestrians, cyclists and other non-motorised users of the roads», and it is the driver's insurance that is meant to do the covering. Several American no-fault statutes go further and route a first-party medical benefit to a struck non-occupant: Kansas requires the policy to include those benefits for «other persons struck by such motor vehicle and suffering bodily injury while not an occupant of a motor vehicle», and North Dakota requires the insurer of a secured vehicle to pay benefits for «accidental bodily injury sustained by any pedestrian in this state as a result of being struck by the secured motor vehicle». In all of those, the policy that answers is somebody else's. Whether your own household already holds a policy that would also answer is a separate question, and Minnesota's priority section is written precisely because both can be true at once.

The vehicle drove off and nobody got the registration. Is that the end of it?

Not as a matter of design, though what exists varies by country. The directive requires each Member State to «set up or authorise a body» to compensate for damage caused by an unidentified vehicle or one that was not insured, adds that the victim «may in any event apply directly to the body», and forbids Member States from letting that body make payment «conditional on the victim establishing in any way that the person liable is unable or refuses to pay». Spain's implementation names the Consorcio de Compensación de Seguros and gives it the job of compensating personal injury «en aquellos casos en que el vehículo a motor causante sea desconocido». The United States has no single national equivalent; some no-fault states run an assigned claims plan instead, and North Dakota's exists for an injured person to whom benefits «are not applicable to the injury» for a reason other than the statute's own exclusions.

Does being on a bicycle change anything compared with being on foot?

In the instruments, less than you would expect; in the tools, more. The directive names pedestrians and cyclists in the same clause. The American no-fault provisions quoted here are drafted around being «struck» while «not an occupant of a motor vehicle», which is a description of position rather than of mode — though North Dakota's own wording says «pedestrian», and whether a cyclist falls inside a given state's word is a question for that state's text rather than for this article. Where the difference is real and visible is the crash simulator: for a struck pedestrian it computes an average fatality risk against impact speed from a published curve, and for a cyclist it reports the impact speed and stops there, because no openly readable cyclist speed–risk curve could be located. The tool says so itself rather than filling the gap.