EVIDENCE LIBRARY
Can a Parking Collision Alone Prove Alcohol Impairment?
A real-car experiment linked increasing blood alcohol with more parking collisions, but single collisions occurred while sober too. The findings do not validate a diagnostic rule for an individual driver.
No. One parking collision is not specific evidence of alcohol impairment. In a 29-driver real-car experiment, single collisions occurred during sober and alcohol-exposed drives. Increasing blood alcohol was associated with a higher collision count, and repeated collisions occurred only after drinking in this sample, but the study did not validate a rule that can diagnose impairment from collision count alone.
What the experiment involved
Tank and colleagues studied 29 licensed drivers, including 13 men and 16 women, on a closed private course. Each completed three drives, with forward, reverse and parallel parking maneuvers. Twenty-five participants drank alcohol after their initial sober drive; four remained sober throughout as controls.
The investigators calculated doses intended to reach 1.10 g/kg blood alcohol. The second drive followed drinking, and the third occurred approximately two hours later without more alcohol. Blood was sampled after each drive. Actual concentrations varied considerably.
The reported unit was grams of ethanol per kilogram of blood. These values should not be silently relabeled as g/L, g/dL or a breath concentration. A calculated dose target also differs from the subsequently measured result.
Cars had manual transmissions, no parking aids and a mechanical speed restriction of approximately 20 km/h. Participants used the same car across their drives and had time to become familiar with the course.
Collision count and a single collision answered different questions
The study comprised 87 drives from 29 people, not 87 independent drivers. At least one collision occurred in 20 drives. Seven drivers collided during their first sober drive, seven during the second drive, and six during the third.
The statistical model, accounting for driver differences and drive number, found an association between increasing BAC and the number of collisions (p = 0.04). Five drives involved more than one collision, and these occurred only in alcohol-exposed drivers, at measured concentrations of at least 0.63 g/kg.
The lowest concentration observed among those five drives is not a validated threshold. The sample was too small to establish that repeated parking impacts cannot occur sober, that every driver above that value will collide, or that two collisions prove alcohol caused an event.
Most collision-involved drives included parallel parking. The findings therefore depend on the maneuvers, course geometry and vehicles used.
Other parking outcomes were not significantly associated with BAC
The investigators found no statistically significant BAC association with completion time (p = 0.41), number of corrective movements (p = 0.25) or final parking position (p = 0.75).
Those null findings do not cancel the collision-count result. They show that different measures of the same general activity can behave differently. A person may finish in a satisfactory position or use an ordinary number of corrections while still having made contact along the way.
The site’s simulator study comparing driving errors and collisions illustrates the same need to specify the endpoint, using a different design and different results.
A collision does not establish whether the driver noticed it
Pressure-sensitive films were attached at likely impact locations. Fifteen films could be evaluated, but some represented multiple impacts because the film was replaced after each drive rather than after every collision.
Some impacts were noticed by both driver and passenger, some by the passenger alone, and some by neither. The investigators could not perform a reliable statistical analysis of impact intensity because the available sample was small. Their observations do not establish an alcohol-specific rule about awareness.
Whether a particular impact could be seen, heard or felt requires its own reconstruction, including the vehicle, impact location and surrounding conditions. Damage alone does not reveal the driver’s subjective experience.
Limits on a DUI opinion
The sober control group contained only four people. Participants knew they were being tested and were motivated to perform well. A learning effect occurred, and the setup differed slightly between the two testing days. The study did not reproduce every roadway or modern parking-assistance system.
The useful conclusion is that a single parking collision is nonspecific, while repeated impacts deserve investigation alongside other evidence. Neither pattern independently supplies the driver’s BAC, establishes causation or determines a legal violation.
Any case opinion should distinguish the measured alcohol result from a behavioral inference and identify the timing of both. The site’s retrograde BAC guide addresses the separate assumptions involved in estimating an earlier concentration.
Primary source
Tank A, Tietz T, Loskant J, Zube K, Ritz-Timme S, Hartung B. Parking behaviour under the influence of alcohol. International Journal of Legal Medicine. 2023;137:1735–1741. https://doi.org/10.1007/s00414-023-03082-2.