This study examined the predictive validity of pre-university examinations test scores (university matriculation examination (UME), Post-UME and pre-degree) for undergraduate academic achievement. The study is planned along the lines of correlational and ex-post-facto research design. A sample of four hundred university science based undergraduates admitted 2006/2007 session which was made up of 100 students selected from each of the four purposively selected universities from south west Nigeria using stratified sampling technique participated in the study. An inventory titled "mode of entry characteristics and academic achievement proforma" was used to collect data from records. The data were subjected to Pearson Product Moment Correlation and multiple regression (Backward solution) analysis. Result reveled that as much as 17.1% of the variations in CGPA is explained by the linear combination of three predictor variables, while pre-degree scores indicated significant contribution to the prediction, UME showed negative contribution to the prediction. The result also showed that significant relationship exists between the independent variables (Pre degree, Post UME scores) and CGPA while there was low relationship between UME and CGPA. Based on the findings of the study, there is need to further strengthen and uphold the pre-degree programme and post-UME tests for continuous utilization as means of recruiting viable candidates for university admission purpose.