OBJECTIVE: Identifying moderators of the effects of self-efficacy-enhancing interventions could improve their efficiency. We examined the effects of a home-based variant of the Chronic Disease Self-Management Program on self-efficacy, and explored the moderating effects of perceived control over self-management (PCSM). METHODS: In a randomized controlled trial, patients (N= 415) aged>40 years with various chronic conditions plus basic activity impairment and/or significant depressive symptoms were randomized to one of three groups: intervention provided in homes or by telephone, v. usual care control. We used mixed effects linear models for repeated measures to examine effects on self-management self-efficacy at 6-month follow-up and explore moderation by PCSM. RESULTS: Only the home intervention had a significant self-efficacy-enhancing effect (Wald test, chi( 2) = 13.8, p = 0.008; effect size = 0.3). The effect was moderated by PCSM, considered as a continuous [effective in subjects with lower PCSM (Wald test, chi(2) = 13.4, p = 0.009)] or categorical (effective only for subjects in the lowest tercile) variable. CONCLUSIONS: People with lower PCSM appear more likely to experience enhanced self-efficacy from chronic illness self-management training than those with higher PCSM. These findings, although preliminary, suggest that office-based measurement of PCSM might identify those chronically ill patients likely to benefit most from self-management training.