Abstract Daily affective reactivity refers to the within-subject relationship between daily stress and daily mood. Most stress researchers have conceptualized daily affective reactivity as a dependent variable to be predicted by individual difference variables such as personality and psychopathology. In contrast, in our recent research, we have conceptualized daily affective reactivity as an independent variable that can predict depressive symptoms. In this article, we summarize three studies that relied on a daily process methodology and multilevel modeling to assess affective reactivity in the context of daily stressful events. Two of the studies (Cohen, Butler, Gunthert, & Beck, 2005; Gunthert, Cohen, Butler, & Beck, 2005) sampled adult outpatients in cognitive therapy and evaluated the predictive role of daily affective reactivity in treatment outcome (depression reduction). A third study (O'Neill, Cohen, Tolpin, & Gunthert, 2004) evaluated the predictive role of college students' daily affective reactivity in the development of depressive symptoms. We consider the strengths and weaknesses of a daily process methodology for research on depression in both clinical and nonclinical samples.