AbstractThe use of citations in place of full custody arrest has generally been considered a liberal reform. It has been suggested, however, that because such a policy might produce a shift in police officers' criteria for arrest, it could result in an increase in the number of people being brought into the criminal justice system. An interrupted time series design was used to study the effect of a citation policy on arrest rates. The predicted increase was found for only one offense. The implications for the evaluation of citation arrest are discussed.