Understanding the underlying processes influencing college students' academic achievement represents an important goal of educational research. The aim of the present study was to examine the utility of the extended Theory of Planned Behaviour (TPB) and the relative influence of cognitive processes and measures of past behaviour in the prediction of grade intentions and actual grades in the context of higher education. Our results show that perceived behavioural control (PBC), subjective norms (SN) and past grades emerged as significant predictors and were able to explain 41% of the variance in grade intentions, while the influence of cognitive processes (planning and future orientation) was not significant. With respect to the prediction of actual grades, we found that habitual school attendance and past grades explained 26% of the variance in actual grades. Again, the influence of cognitive processes (planning and future orientation) was not significant. The main discussion is centred on the utility of the extended TPB in the prediction of academic achievement and the complex relationship between cognitive processes and past behaviour in predicting human actions in general. Limitations and implications of the study are also discussed.