With the dismantling of apartheid and the opening of universities to people from sociopolitically disadvantaged communities, the issue of relevant selection procedures has become increasingly pressing in South Africa. This study aimed at exploring the relative value of various predictots for disadvantaged versus advantaged students. Subjects comprised the 26 students enrolled in a pre-university Faculty of Commerce bridging yesr, 18 of whom were ftom educationally disadvantaged backgrounds. Generally, the academic success of disadvantaged, was considerably less predictable than that advantaged students. Further, universally-used matriculation criterion was not a significant predictor for either the disadvantaged or the advantaged, while the dynamic and process measures applied were also not significant predictors for disadvantaged students. The results were seen as supporting the need for an alternative approach to selection. and a of enrichment programmes themselves.