We conducted four studies to explore people's self-prediction processes. Study 1 examined the types of information people report using when making self-predictions. Five categories of information were determined. Studies 2 and 3 examined the relation between the use of different categories of information and self-prediction accuracy. Using correlational and experimental methodologies, these two studies demonstrated the utility of attending to personal base rate and personal disposition information in formulating accurate self-predictions. Individual differences in accuracy as a function of public and private self-consciousness were also evident. Study 4 found that more certain self-predictions, as well as predictions that were distinct from what was expected for the average individual, were more accurate. Overall, the findings suggest the importance of attending to individuating information in formulating accurate self-predictions.