Background Proper classification of child occupant restraint use is dependent on the age of the child occupant. Observations of vehicle restraint use involve estimating child age. If estimates of age are incorrect, then a potential for misclassification of restraint use exists.

Objective To compare estimated and confirmed child occupant age and calculate the impact of errors in age estimates on the proportion of children classified as properly restrained.

Methods Observations of restraint use were completed for occupants 0–8 years of age at two health clinics. After initial observation, we approached the driver to confirm the child's age. Each child's restraint use was classified as either compliant or not compliant with state law, based on type of restraint used and based on the child's estimated and confirmed ages.

Results Classification of age categories for child occupants (n=218) was correct in 86.3% of observations. For 48.6%, the confirmed and estimated age matched exactly, and for 98.1%, age matched within ±1 year. Overall, compliant restraint use based on estimated age was 39.4%, and based on confirmed age was 38.5%. In paired comparisons, restraint use based on estimated age versus confirmed age was concordant for more than 95% of children.

Conclusions The level of accuracy for age estimates was sufficient for making estimates of compliant restraint use. Errors in estimated age resulted in a less than 1 percentage point difference in overall proper restraint use calculations. The results suggest that such observations can be a reliable measure of proper child occupant restraint use.