An observational study of nine roadway segments in three towns in Connecticut was conducted to compare the race of motorists observed on the road to the race of motorists stopped by the police in those towns. The results showed that there is a very close relationship between police stops and motorists observed on the road. In six of the nine locations there was a non-significant difference between the race of the motorists observed and the race of those stopped. In two locations black motorists were stopped less frequently than all other races, and in one location black motorists were stopped more frequently than all other races. The study argues that census based estimates for driving population are not valid and should be eliminated from evaluating police behavior.