Deterrence theory implies that a strong response by a juvenile officer to delinquent behavior will reduce the probability of the behavior recurring; however, labeling theory suggests that a strong response will increase the probability of future delinquency. Two empirical questions must be answered in establishing support for either deterrence or labeling theory in the police handling of juveniles; the effect of the arrest on recidivism and reasons for the effects found. There has been very little research on the reasons for observed patterns of recidivism; consequently, there is no strong evidence for the validity of either theory as applied to police-juvenile encounters. Some helpful conclusions can be drawn from the research, however. Several factors could mask any suppression of delinquent behavior. Age and number of prior offenses are positively correlated with recidivism. To a degree unaccounted for by age and prior offenses, more severe dispositions are given to juveniles more likely to recidivate; therefore, comparisons between juveniles who have received different treatment are biased against whatever suppression effects exist. The second conclusion is that there are individual differences between juveniles related to recidivism, and they exist prior to any interaction with police. The third conclusion is that police may unknowingly minimize the suppressive effect of arrest in an attempt to avoid labeling effects, and failing to follow arrest with significant court action may also undermine any arrest effectiveness. The final conclusion is that current suppression effects are not great enough to overcome the factors that mask them. These conclusions suggest directions for further research. Thirty-five references are listed. For related material, see NCJ 89520 and NCJ 89522-25.