Drawing on a conceptual model integrating research on training, work–family interventions, and social support, we conducted a quasi-experimental field study to assess the impact of a supervisor training and self-monitoring intervention designed to increase supervisors' use of family-supportive supervisor behaviors. Pre- and postintervention surveys were completed, 9 months apart, by 239 employees at 6 intervention (N = 117) and 6 control (N = 122) grocery store sites. Thirty-nine supervisors in the 6 intervention sites received the training consisting of 1 hr of self-paced computer-based training, 1 hr of face-to-face group training, followed by instructions for behavioral self-monitoring (recording the frequency of supportive behaviors) to facilitate on-the-job transfer. Results demonstrated a disordinal interaction for the effect of training and family-to-work conflict on employee job satisfaction, turnover intentions, and physical health. In particular, for these outcomes, positive training effects were observed for employees with high family-to-work conflict, whereas negative training effects were observed for employees with low family-to-work conflict. These moderation effects were mediated by the interactive effect of training and family-to-work conflict on employee perceptions of family-supportive supervisor behaviors. Implications of our findings for future work–family intervention development and evaluation are discussed.