OBJECTIVE: To explore whether adding a gender and HIV training programme to microfinance initiatives can lead to health and social benefits beyond those achieved by microfinance alone. METHODS: Cross-sectional data were derived from three randomly selected matched clusters in rural South Africa: (i) four villages with 2-year exposure to the Intervention with Microfinance for AIDS and Gender Equity (IMAGE), a combined microfinance-health training intervention; (ii) four villages with 2-year exposure to microfinance services alone; and (iii) four control villages not targeted by any intervention. Adjusted risk ratios (aRRs) employing village-level summaries compared associations between groups in relation to indicators of economic well-being, empowerment, intimate partner violence (IPV) and HIV risk behaviour. The magnitude and consistency of aRRs allowed for an estimate of incremental effects. FINDINGS: A total of 1409 participants were enrolled, all female, with a median age of 45. After 2 years, both the microfinance-only group and the IMAGE group showed economic improvements relative to the control group. However, only the IMAGE group demonstrated consistent associations across all domains with regard to women's empowerment, intimate partner violence and HIV risk behaviour. CONCLUSION: The addition of a training component to group-based microfinance programmes may be critical for achieving broader health benefits. Donor agencies should encourage intersectoral partnerships that can foster synergy and broaden the health and social effects of economic interventions such as microfinance.