Drawing from the meta-construct model of cognition (Ingram, 1984; Ingram & Kendall, 1986), the goal of the present study was to examine whether the structural (i.e., self-complexity; SC, Linville, 1985) and propositional components of schemas (dysfunctional attitudes; DAS, Weissman & Beck, 1978), independently and in interaction with each other and stressors, lead to changes in depressive symptoms. The prediction was that if negative self-attributes across different self-aspects in a specific domain, interpersonal or achievement, are highly distinct (i.e., high negative SC) or if positive self-attributes across different aspects of self are redundant (i.e., low positive SC), then the DAS would be more likely to lead to higher levels of depression when domain-congruent stressors occur. To test the main effect, two-way interaction, and three-way interaction hypotheses, the present study used a two month longitudinal design involving three assessment periods, separated by one month. At the baseline, a total of 189 students participated in the study. Of these students, 163 and 121 students participated in the first and second follow-ups, respectively. Multiple hierarchical regression analyses were used to examine two-way and three-way interactions. Results showed strong support for the predictive power of negative SC with respect to depressive symptomatology. The DAS, on the contrary, was a concurrent factor related to depressive symptoms. Further, the present study did not provide supportive evidence for the diathesis-stress model of depression. Although contradicting expectations, the pattern of relationships between interpersonal negative SC, DAS, distal stressors suggested promising venues for future research.