What a crash reconstructionist can determine that a police report may not
What reconstruction can show about speed and impact, and when families may need it
Safety Behind the Wheel Foundation built this guide from National Highway Traffic Safety Administration (NHTSA) investigation programs and guidance, federal event data recorder rules, the Federal Rules of Evidence and published research on collision data and reconstruction methods. Every source is listed at the end.
What a police report is built to do
Officers at an accident scene have many jobs at once. They get help for the injured, secure the scene, clear traffic and record the basic facts. The report they file is shaped by that moment.
All 50 states, the District of Columbia and several territories collect crash data through police reports [1]. NHTSA publishes a voluntary guideline, the Model Minimum Uniform Crash Criteria (MMUCC), that sets a minimum, standard list of facts to record [1]. Its newest edition came out in 2024 [1]. Even so, states differ a great deal in how they report crash data [1].
NHTSA estimates that 6 to 7 million police reported crashes happen each year [2]. Officers usually arrive after the vehicles have stopped, so many details come from what drivers and witnesses say.
Where reports can fall short
A review of crash data quality found that the most serious problems included wrong locations and times, misjudged injury severity, and errors in naming collision causes [3]. In one South Korean study, researchers compared police records of taxi and pedestrian collisions with video from in vehicle recorders [4]. The police records relied on statements from the people involved and from witnesses [4]. On average, the recorded location was off by about 84 meters (about 277 feet). The time was off by about 29 minutes, and the speed by about 9 km/h, or roughly 5.6 mph [4].
None of this means officers do poor work. It means a police report is a starting record, not an engineering study. If your report has a plain factual error, the Drive With Care guide to crash reports and evidence explains how to ask for a correction.
What reconstruction adds
Crash reconstruction uses physics, engineering and careful evidence review to work out how a collision happened. Where was each vehicle before impact? How fast was it going? What did each driver do in the last few seconds? How did the people inside move, and what did they strike?
The goal is to test every version of events against the physical evidence. A driver may remember going 30 mph. The tire marks, crush damage and recorded data either fit that memory or they do not.
NHTSA’s research programs show what deeper investigation looks like. In its Crash Investigation Sampling System, trained crash technicians record what the scene shows, including skid marks, spilled fluids and objects the vehicles struck [5]. They inspect the vehicles, record the damage and identify the parts inside that occupants hit [5]. Then they hold private interviews and review medical records [5].
NHTSA’s Special Crash Investigations program has done the agency’s most detailed crash work since 1972 [6]. It gathers hundreds of details on each case, from the vehicles and roadway to how people were hurt and how safety systems worked [6]. Each year it selects more than 100 collisions for study [6]. Next to millions of police reported collisions [2], that is a tiny share.
A private reconstruction uses many of the same tools. For a serious collision, also ask the police agency whether a specialized crash team worked the scene and whether its findings are in a separate report.
The evidence a reconstructionist pulls together
No single piece of evidence tells the whole story. A reconstruction is strongest when several independent sources point to the same answer.
Comparison compiled by Safety Behind the Wheel Foundation
Evidence used in a reconstruction and its limits
| Evidence | What it can help show | Common limits |
|---|---|---|
| Tire marks, gouges, debris, final resting spots | Paths of travel, point of impact, braking or skidding [5] | Fade with rain, traffic and cleanup |
| Crush depth and damage pattern | Crash severity, known as delta V [9] | Estimates carry error [9] |
| Event data recorder | Speed, braking, throttle, belt use, air bag timing [8] | Records only a few seconds [7] |
| Video and photos | Timing, speed, signal color, lighting [10] | Camera angle, clock errors, lens distortion |
| Marks inside the vehicle, medical records | Seating, belt use, how injuries happened [5] | Needs an early, hands on inspection |
| Road layout, signs and sight lines | What each driver could see, and when | Conditions may change after the collision |
Delta V, a measure of impact force
Delta V is the change in a vehicle’s speed and direction during the impact itself. Engineers treat it as a gauge of how severe a collision was and how likely injury is [9]. A reconstructionist can estimate it from how deeply the vehicle was crushed and how stiff that model is [9]. Event data recorders may also log it directly [8].
Event data recorders
An event data recorder (EDR) saves technical vehicle and occupant information for a brief time around a crash, which NHTSA describes as seconds, not minutes [7]. Federal rules cover light vehicles built on or after September 1, 2012, that have an EDR [8]. These devices must capture items such as speed, accelerator and brake use in the 5 seconds before a crash, driver belt use and frontal air bag timing [8]. A companion Drive With Care guide covers EDR data in depth.
Video
Surveillance and dashcam video can show timing, speed and signal color. Researchers have built methods to estimate speed from crash video frame by frame, even on curves [10]. Video still needs checking, because camera clocks can be wrong and wide lenses can stretch distances. If you have dashcam footage, see how to save your dashcam video after a crash.
Questions a reconstruction can help answer
A good reconstruction tests specific questions with measurements, such as:
- How fast was each vehicle going when braking began and at impact?
- In a chain reaction collision, which impact came first?
- Did a driver brake, steer or speed up in the final seconds?
- Who crossed into whose path?
- Given the lighting, curves and objects in the way, could a driver have seen the hazard in time to stop?
- Who was driving, and was each person wearing a seat belt?
- Were the impact forces consistent with the injuries people suffered?
That last question often comes up when an insurer says a collision was too minor to hurt anyone. Biomechanical engineers, who study how forces affect the body, can feed crash data from a reconstruction or an EDR into computer models of a seated person. One peer reviewed study used this method for low speed to moderate speed frontal collisions [11]. Its model matched crash tests closely for some head measures but less well for others [11]. This kind of analysis does not diagnose anyone. It can test whether a “too minor to matter” claim holds up.
When a safety system’s performance is in question
Families often ask why an air bag stayed folded, or why a car with automatic emergency braking did not stop. These are fair questions. A reconstruction can help separate a system working within its design limits from a possible defect.
Air bags
NHTSA says frontal air bags are generally designed to deploy in moderate to severe frontal or near frontal collisions [12]. That is roughly like hitting a solid wall at 8 to 14 mph or more [12]. They are not meant to deploy in every crash [12]. Many advanced systems also switch off the front passenger air bag when they sense a child, a small adult or an empty seat [12]. So when an air bag does not deploy, that may be normal, or it may not. An engineer compares the impact direction and force with the system’s design, using EDR records of air bag timing and belt use [8].
Driver assistance features
Level 2 driver assistance can steer and control speed at the same time while the driver supervises. If it was in use within 30 seconds of a crash, NHTSA’s Standing General Order may require a report to the agency [13]. That applies when the crash involves a death, a hospital trip, an air bag deployment or a struck pedestrian, cyclist or other vulnerable road user [13]. The companies named in the order file the reports, and NHTSA posts the data from time to time [13]. The order was most recently amended in 2025 [13].
Reporting a suspected defect
If you believe a safety feature failed, you can file a complaint with NHTSA online or through its Vehicle Safety Hotline at 888-327-4236 [14]. Safety Behind the Wheel Foundation encourages families to report any safety feature they believe failed. NHTSA adds complaints to a public database after removing personal details, and it can open investigations based on them [14]. A recall follows when the maker or NHTSA decides a vehicle poses an unreasonable safety risk [14].
When families might need a reconstruction
Not every collision needs one. After a minor wreck with clear fault, a good personal record and the police report are usually enough. A reconstruction is worth discussing with an attorney or insurer when:
- Drivers or witnesses tell stories that cannot all be true.
- Someone died or was badly hurt, and the person who could explain what happened cannot speak.
- The report’s conclusion rests mainly on one person’s statement.
- An insurer says the damage was too light to cause the injuries.
- An air bag, seat belt, brake or driver assistance feature seemed to fail.
- A truck, bus or work vehicle was involved and may carry its own recorders.
- A hit and run driver left only debris and damage behind.
Most families do not hire a reconstructionist on their own. More often, an attorney arranges one, and insurers or prosecutors may hire their own experts. If the other driver’s insurer denies fault, see when the other driver’s insurer says its driver wasn’t at fault.
Act early, because evidence does not wait
Physical evidence starts to disappear right away. Tire marks wear off and debris is swept up. Vehicles get repaired, sold for salvage or crushed. Video systems often record over old footage.
These steps help keep your options open while you decide:
- Write down your own account, keeping what you saw apart from what others told you.
- Tell your insurer, and your attorney if you have one, that the vehicle may be needed as evidence.
- Ask that no one repair, sell or crush the vehicle until everyone involved agrees it can be released.
- Find out where each vehicle is stored and who controls it, since storage fees may apply.
- Photograph the vehicle inside and out without moving parts or wiring.
- Ask nearby homes and businesses about video right away.
- Leave vehicle data downloads to a trained person with the right tool and proper permission.
- Keep a simple log of who handles the vehicle, when, and why.
A companion Drive With Care guide explains how to preserve a badly damaged vehicle in more detail.
How to choose and work with a reconstructionist
A reconstruction is only as good as its data and methods. Courts use a similar test. Under the federal rule on expert witnesses, an expert’s opinion must rest on enough facts or data and on reliable methods, reliably applied to the case [15]. A 2023 change stressed that the side offering the expert must show these conditions are more likely than not met [15]. State courts have their own rules, but this is a useful yardstick for any report you read.
Expect ranges, not perfect numbers. A careful report might say a car was going between 38 and 44 mph rather than exactly 41. Be wary of anyone who offers certainty without showing calculations and data. Ask for a plain language summary along with the technical report.
What reconstruction cannot tell you
Reconstruction is powerful, but it has limits.
- Every method has error. One study checked 478 vehicles in frontal collisions against their EDR data [9]. A damage based method used in federal crash studies underestimated delta V by 13% on average, down from 23% in an older version [9]. Good engineers crosscheck one method against another.
- Recorded data is brief. EDRs save only seconds of data and only certain items [7, 8]. They do not show what happened earlier in the trip.
- Physics shows what happened, not why. Data can show that a driver never braked. On its own, it rarely proves whether the driver was distracted, tired or impaired.
- Missing evidence limits answers. If a vehicle was crushed or a scene was never measured, some questions may stay open.
- Experts can disagree. Differences usually trace to different assumptions, so ask each expert to explain theirs.
- It is not a verdict. Insurers, courts and state law decide fault and payment. See fault, settlements, deadlines and your legal rights.
Reading a reconstruction of a collision that hurt you or took someone you love can be very hard. It is fine to ask for a short summary first or to read it with a trusted person beside you.
Frequently asked questions
Is an accident reconstructionist the same as the officer who wrote the report?
Usually not. The officer who writes a standard report records the scene, the people involved and their statements. A reconstructionist does a deeper engineering analysis, often days or weeks later, using measurements, vehicle inspections and recorded data. Both records can be useful, and they often answer different questions.
Can a reconstruction change the police report?
Not directly. The police agency decides what its report says. You can share new findings and ask the agency to review them, but it may keep its original conclusions. Insurers and courts can weigh a reconstruction alongside the police report, so a sound analysis can matter even if the report never changes.
Does my car have an event data recorder?
It may. Federal rules set what the device must record in light vehicles built on or after September 1, 2012, that have one [8]. To learn about your vehicle, check the owner’s manual or ask the dealer. Do not try to read or reset the data yourself. Leave any download to a trained person with the proper tool and legal permission.
Why didn’t my air bag deploy?
Frontal air bags are generally designed for moderate to severe frontal or near frontal collisions, not every crash [12]. Some systems also turn off the passenger air bag for a child, a small adult or an empty seat [12]. If the air bag did not deploy, that may be normal for that crash. If the impact direction and force suggest it should have deployed, an engineer can review the evidence, and you can file a complaint with NHTSA [14].
Who pays for a crash reconstruction?
It depends on who hires the expert. Insurers and prosecutors may pay for their own. When a family’s attorney hires one, ask how the cost will be handled and get it in writing. Before work begins, ask for an estimate that lists what it covers, such as inspections, data downloads, reports and testimony.
The bottom line
Recommendation from Safety Behind the Wheel Foundation
Safety Behind the Wheel Foundation encourages families facing a serious, disputed collision to ask early whether a reconstruction is needed and to protect the evidence while they decide. A police crash report is a valuable first record, but it is built for speed and consistency, not engineering detail. When the facts of a serious collision are in dispute, a reconstruction can test each account against tire marks, crush damage, recorded data, video and injury patterns.
Not every family will need one. If you do, act early to protect the evidence, choose an expert who shows their work, and expect ranges instead of perfect numbers. Clear answers about what happened can support fair claims, safer vehicles and, for many families, a measure of peace.
Related articles
- Crash reports and evidence: keeping a reliable record after a serious crash
- How to save your dashcam video after a crash
- When the other driver’s insurer says its driver wasn’t at fault
- After a serious crash: fault, settlements, deadlines and your legal rights
- Understanding traffic safety data: how to read crash statistics wisely
Sources and further reading
- 1National Highway Traffic Safety Administration. Model Minimum Uniform Crash Criteria. Accessed October 2026 (MMUCC 6th Edition, 2024). nhtsa.gov/mmucc-1
- 2National Highway Traffic Safety Administration. Crash Report Sampling System. Accessed October 2026. nhtsa.gov/…/crash-report-sampling-system
- 3Imprialou M, Quddus M. Crash data quality for road safety research: Current state and future directions. Accident Analysis and Prevention, 2017 (via PubMed). pubmed.ncbi.nlm.nih.gov/28262098
- 4Chung Y, Chang I. How accurate is accident data in road safety research? An application of vehicle black box data regarding pedestrian-to-taxi accidents in Korea. Accident Analysis and Prevention, 2015 (via PubMed). Data year: crash records analyzed in the study (South Korea; year range not stated in abstract). pubmed.ncbi.nlm.nih.gov/26298271
- 5National Highway Traffic Safety Administration. Crash Investigation Sampling System. Accessed October 2026. nhtsa.gov/…/crash-investigation-sampling-system
- 6National Highway Traffic Safety Administration. Special Crash Investigations. Accessed October 2026. nhtsa.gov/…/special-crash-investigations
- 7National Highway Traffic Safety Administration. Event Data Recorder. Accessed October 2026. nhtsa.gov/research-data/event-data-recorder
- 8Electronic Code of Federal Regulations. 49 CFR Part 563, Event Data Recorders. Up to date as of October 1, 2026. ecfr.gov/…/part-563
- 9Hampton CE, Gabler HC. Evaluation of the Accuracy of NASS/CDS Delta-V Estimates from the Enhanced WinSmash Algorithm. Annals of Advances in Automotive Medicine, 2010 (via PubMed). Data year: 2000 to 2008 crash cases. pubmed.ncbi.nlm.nih.gov/21050607
- 10Choi Y, et al. Cross-ratio and vehicle dynamics-based speed estimation for traffic accident analysis. Forensic Science International, 2025 (via PubMed). pubmed.ncbi.nlm.nih.gov/41066911
- 11Jeffs SB, Nolasco LA, Petroskey KJ. Predicting head injury metrics during low- to moderate-speed frontal collisions using computational simulations. Accident Analysis and Prevention, 2023 (via PubMed). pubmed.ncbi.nlm.nih.gov/37769479
- 12National Highway Traffic Safety Administration. Vehicle Air Bags and Injury Prevention. Accessed October 2026. nhtsa.gov/vehicle-safety/air-bags
- 13National Highway Traffic Safety Administration. Standing General Order on Crash Reporting. Updated August 2026 (third amendment effective June 16, 2025). nhtsa.gov/…/standing-general-order-crash-reporting
- 14National Highway Traffic Safety Administration. Check for Recalls: Vehicle, Car Seat, Tire, Equipment. Accessed October 2026. nhtsa.gov/recalls
- 15Legal Information Institute, Cornell Law School. Rule 702. Testimony by Expert Witnesses. Federal Rules of Evidence, as amended effective December 1, 2023. law.cornell.edu/rules/fre/rule_702
This article reflects the safety priorities of fair, evidence based answers after a serious collision. It is general education, not legal, medical, insurance or financial advice. Investigation practices, court rules and federal reporting orders can change, so verify details with official sources. For your own situation, talk with a licensed attorney, your insurer or the police agency that investigated. To report a suspected vehicle safety defect, contact NHTSA. Reviewed October 2026.
Cite this article: Safety Behind the Wheel Foundation. “What a crash reconstructionist can determine that a police report may not.” drivewithcare.org, https://drivewithcare.org/crash-reconstruction-beyond-police-report/. Reviewed October 2026.
Special thanks to our contributors
We appreciate the professionals, experts, and community members who share their knowledge and perspectives to help make this resource better. Your insights are always welcome and appreciated.
Have professional expertise, firsthand experience, or a perspective that could make this resource better?
We welcome your input. Your knowledge and experience can help us make this resource more useful, accurate, and meaningful to the people who rely on it.
Thank you. Your contribution is on its way.
We’re grateful you took the time to share your knowledge with us. Our team reviews every contribution with care. If it helps improve this resource, we’ll update the page and recognize you the way you chose.
Keep learning
Safe Driving Resource Center
Guides, self-checks and articles for drivers, parents of new drivers, and families recovering after a crash.
