Stigma is considered an important barrier to seeking mental health services. Two types of stigma exist: public stigma and self-stigma. Theoretically, it has been argued that public stigma leads to the development of self-stigma. However, the empirical support for this assertion is limited to cross-sectional data. Therefore, the goal of this research was to examine the relationship between public stigma and self-stigma over time. Perceptions of public and self-stigma were measured at Time 1 (T1) and then again 3 months later at Time 2 (T2). Using structural equation modeling, we conducted a cross-lag analysis of public stigma and self-stigma among a sample of 448 college students. Consistent with assertions that public stigma leads to the development of self-stigma, we found that public stigma at T1 predicted self-stigma at T2, whereas the converse was not true. These findings suggest that if self-stigma develops from public stigma, interventions could be developed to interrupt this process at the individual level and reduce or eliminate self-stigma despite perceptions of public stigma. (Contains 1 table, 2 figures and 2 footnotes.)