With a pretest-posttest matched control group design, attitudinal change toward themselves and the prison staff was assessed through two semantic differential measures. A multivariate analysis of variance indicated no significant differences in attitude change between the thirty-three volunteer treatment group wards and the the thirty-four no-treatment group wards. Therefore, it was concluded that group therapy treatment had no rehabilitative advantage over vocational training. However, the analysis did reveal that wards who were convicted of crimes against persons reported more positive attitudinal change toward themselves and prison authority within a three-month time period than did wards convicted of crimes against property. (Author abstract)