The studies aim is to see if there is a direct relationship between levels of family/social support and levels self-perceived hostility. The main hypothesis is that levels of hostility will be related to levels of social support. Two questionnaires used in this study were the Buss-Durkee Hostility inventory (1983) and a family/social support scale which was modelled on the Family Support Scale by Dunst, Jenkins and Trivette (1984). The number of participants involved was 20 consisting of 40% convicted and 60% non-convicted. The predictor variable is Family/Social support, the Criterion Variable is self-perceived Hostility. Age was used as the matching criterion for the two groups of participants. The experimental group were the young offenders and the control group were the students from the secondary school system. The variables were a between Subjects design. The control group study was conducted in their classroom. The experimental group filled in their questionnaires in a room with one of their care assistants. The study concluded that there were no significant results. The null hypothesis was accepted.