Processing speed and response control are fundamental properties of brain function and potential markers of cognitive ability. This study, a secondary analysis of a randomized controlled trial, examined whether eye-tracking measures of saccadic reaction time and gaze control are associated with an established cognitive ability test, Raven’s coloured progressive matrices (CPM), among 13-year-old rural Malawian adolescents (1003 participants, 50.3% boys). Mean prosaccadic reaction time (_pSRT_m), antisaccade error rate (PE) and CPM result were obtained for 760 (75.8%), 621 (61.6%) and 997 (99.4%) children. Pearson correlation and linear regression were used to evaluate the association of the tasks.Faster _pSRT_m and lower PE were very weakly associated with higher CPM score (rs -0.12, p = .001 and -0.11, p = .006). In the covariate adjusted regression models, faster prosaccadic reaction time (pSRTm) was very weakly associated with higher scores in CPM test (adjusted coef -0.02, 95%CI (-.03- -.002), p = .03), but antisaccadic errors were not associated with CPM score (adjusted coef -0.63, 95%CI (-1.60 -.33), p = .20). Post hoc-analyses suggested stronger associations between eye-tracking measures and CPM among participants with more schooling (years in school <4.5 or >4.5, rs between _pSRT_m and CPM -0.05 and -0.21; between PE and CPM -0.01 and -0.39). The results confirm the predicted association between saccadic speed and cognitive ability in an understudied population, but the connection is weaker than expected according to earlier studies. Schooling potentially moderates the association between eye-tracking tests and CPM.