The study measured recidivism by subsequent convictions. It tested two hypotheses, one related to the incapacitation effect of imprisonment and the other related to the deterrent effect of imprisonment. The two groups of participants included: (1) 779 offenders who, under the State's sentencing guidelines, should have gone to prison but did not due to judicial discretion; and (2) 773 offenders who, under the guidelines, should have gone to prison and did. The 4-year follow-up examined the recidivism rate of each group by ethnicity, age, type of crime (misdemeanor or felony), offense gravity score, and prior offense seriousness. Preliminary findings indicated that the hypothesized effects of incarceration generally did not exist for this group of offenders. However, the two groups were not equivalent comparison groups. In addition, some of the incarcerated offenders served less time than others and thus became at risk for a greater part of the time period. Thus, findings were inconclusive. Unless researchers determine and society changes the causes of what makes some individuals offend, the reactive approaches currently used will continue and the crime rate will not improve. Tables and 34 references