Car, truck, and other vehicle accidents often result whiplash injuries. Whiplash injuries occur at various speeds and even low impact collisions have shown to cause whiplash and other soft tissue injury. What usually happens is that soft tissue is stretched beyond its elastic limitations, meaning it reaches a point when it loses its elasticity and cannot rebound to its normal position.

Whiplash injury occurs in car accidents when a person's vehicle is applied force usually, by another vehicle. The head is pushed in one direction and then when the vehicle is stopped the head is pushed in the opposite direction, this is what causes the soft tissue around the neck to stretch, often beyond its elastic limitations, this is often a source of not only neck pain, but also headaches.

Seat belts have been effective in saving lives, but there are also studies which indicate that the very same seat belts increase the risk of whiplash injury or injury to the neck. The seat belts keep the body and place, but the head is left with no protection and the rotation and forces applied to the head cause strain and injury to the neck. In at least one study it was determined that accident victims were more likely to suffer cervical injury (neck) when they were wearing seatbelts by a factor of 1.58 to 1.

Insurance companies often rely on engineers to testify in court concerning the forces applied as a determining factor as to whether or not there is injury. More often than not there is a reference to Delta-V, but studies have shown that injury cannot be predicted solely by knowing the delta-V of a collision. Even under controlled conditions the delta-V was insufficient to determine whether an injury would occur or not. Delta V is simply an equation taken from physics showing a change in velocity.

The typical crash analyses involves an engineer looking at pictures of the crashed car and then looking at a repair estimate to determine how it compares to similar cars or the same model car that have been crashed tested. The biggest problem with this analyses is that it is comparing apples to oranges. The crash tested vehicles are generally tested against a solid wall, where the actual crash for the injury claimant has crashed with another vehicle. The tested crash is under control conditions and the crash is done with great care. The real world crash has no such conditions. The timing in the crash is different for each crash the location of the damage varies. The angles of the vehicles at the time of the collision, the seating position of the driver and passengers, and the direction where the persons face varies in every single accident.

Injury cannot be predicted for any one individual because individuals vary greatly. Everyone has a different posture, different tensile strength of the ligaments, a different position in the vehicle at the time of the collision, different spinal canals, a different nervous system, and reaction to stimulus. Any one person can also change from month to month, maybe one started exercising three weeks ago, or changed the diet substantially, or had another injury, or over stressed the body.

Insurance companies nonetheless relay on these engineers to testify consistently on claims that low impact collisions cannot cause injury. Accident reconstructionists often use software to calculate speeds of impact by looking at several factors, sometimes just pictures. The software is usually not accurate enough for low impact collisions and the results can vary in extremes from engineer to engineer. The calculated speed for the collision varies greatly depending on how the calculations are entered, most software cautions that it is not reliable for low speeds and studies have shown that these computations are unreliable. The results also vary depending on the relationship between the engineer doing the calculations and the insurance company paying the tens of thousands of dollars to the engineer.

In addition to the engineers, insurance companies also hire medical doctors and chiropractors to render opinions. Opinions which are more likely than not biased.