In this study the authors examined psychosocial variables as mediators for fruit and vegetable (FV) intake in a clustered, randomized effectiveness trial conducted in African American churches. The study sample included 14 churches (8 intervention and 6 control) with 470 participants from the intervention churches and 285 participants from the control churches. The outcome of FV intake and the proposed mediators were measured at baseline and at 6-month follow-up. Structural equation modeling indicated that the intervention had direct effects on social support, self-efficacy, and autonomous motivation; these variables also had direct effects on FV intake. Applying the M. E. Sobel (1982) formula to test significant mediated effects, the authors confirmed that social support and self-efficacy were significant mediators but that autonomous motivation was not. Social support and self-efficacy partially mediated 20.9% of the total effect of the intervention on changes in FV intake. The results support the use of strategies to increase social support and self-efficacy in dietary intervention programs.