Students answered knowledge questions and specified punishment choices for crime scenarios. The subjects were 141 undergraduates at the University of California, Irvine. The treatment group consisted of 65 students enrolled in a course titled "Prisons, Punishment, and Corrections" that had a class enrollment of 127. The comparison group was composed of 76 students enrolled in another criminology class, "White-Collar Crime," offered during the same quarter that had a class enrollment of 170. The primary variables of interest regarding punitiveness and punitiveness change were the pretest punitiveness score and the factual knowledge change in the respondents from time 1 to time 2. Findings show that the corrections course significantly increased the knowledge of the students. As hypothesized, punitiveness within the corrections course sample, as measured by the composite score of the students' punishment choices across vignettes, significantly decreased over time. Both before and after the corrections class, the students preferred that criminals who committed violent crimes be incarcerated. The greatest decrease in punitiveness among the corrections students occurred for nonviolent offenders. Contrary to politicians' view that people want all criminals incarcerated, a majority of students chose less punitive alternatives than jail or prison for all types of nonviolent criminals at both pretest and posttest. Even fewer chose incarceration for nonviolent offenders at time 2 than at time 1. The author discusses study limitations and directions for future research. 4 tables , 9 notes, and 35 references