Despite the rapid growth of health microinsurance (HMI) programs, there is little empirical evidence of their distributional consequences. This paper suggests that the concern of HMI leading to regressive transfers from poor to relatively wealthier households is less important than expected based on an imbalanced randomized control trial with a community based HMI programme in rural Maharashtra in India. Below the Poverty Line (BPL) households consume only 66 per cent of non-BPL households’ annual health care expenditures, suggesting that poorer households subsidize wealthier households. Yet, HMI differentially increases BPL households’ health care expenditure by almost USD 3 per month of coverage. Poorer households are also more sensitive to cost reductions in health care and gains in health care are concentrated among them. These findings suggest that HMI narrows the gap in health care consumption among the insured.