The study investigated the relationship between stress and academic performance among government-sponsored undergraduate students from the University of Nairobi in Kenya. The mediating roles of the students’ age, gender, locus of control, level and course of study in the relationship between stress and academic performance were also examined. The sample consisted of 319 male and 265 female students selected using stratified random sampling techniques, from all the six colleges of the university. The study was carried using a cross-sectional survey design involving levels one to five of the academic programs. Data was collected using questionnaires that measured stress and locus of control. Academic performance was assessed from the students’ academic transcripts. Both descriptive and inferential statistics were used for data analyses. Two-way and three-way chi- square statistics were used to test the statistical significance of the hypothesis.. The analyses were done using the Statistical Package for Social Sciences (SPSS) computer program. Regression analysis was conducted to find out how the confounding variables contributed to the relationship between stress and academic performance. Results showed that most of the students (64.4%) reported that they experienced between moderate to high levels of stress while just over a third (35.6%) reported low stress levels. The relationship between stress and academic performance was statistically significant (χ2=9.49, N=584, df=4, p=0.048). The relationship between stress level and academic performance was significant within 19 to 22 years, 23 to 26 years, males, females, College of Humanities and Social Sciences, College of Agriculture and Veterinary Sciences, levels one and four of study, internal locus of control, and external locus of control. Regression analysis showed that the higher the stress level, the poorer is the academic performance. However, only course/college appears to have statistically significant effect on the relationship between stress and academic performance. The cofounding effect on the stress and academic performance is complex and needs further investigation. The findings indicate the need for relevant authorities to institute programs that will lower the experience and effects of stress among university students. Further research is recommended to investigate the areas where the results were not significant.