Magister Public Health - MPH / The study had an observational, descriptive, cross sectional design. The sample consisted of 205 participants selected through convenience sampling. All students who presented for HIV testing during the colleges’ HIV Counselling and Testing campaign in December 2011 were asked to complete a self-administered closed-ended questionnaire voluntarily. The data was analysed with EPI Info versions 3.5.1. A frequency analysis was conducted for all the variables such as demographics, risk perception to HIV, knowledge of HIV, HIV testing behaviour. The association between HIV knowledge, risk perception and risk behaviours were assessed for gender differences. Chi-square analysis was conducted to determine the association between the main variable of interest HIV risk behaviour and each other explanatory variable (demographics, HIV knowledge, risk perception etc). Results were discussed using the theory of planned behaviour.