This paper contributes to the discussion around ex-post (increased utilisation of health care) and ex-ante (changes in health behaviours) moral hazard in supplemental private health insurance. Applying a range of methodologies to data from the Russian Longitudinal Monitoring Survey—Higher School of Economics we exploit a selection mechanism in the data to compare the impact of workplace provided and individually purchased supplemental health insurance on the utilisation of health care, on a range of health behaviours and on self-assessed health. We find compelling policy-relevant evidence of ex-post moral hazard that confirms a theoretical prediction and empirical regularity found in other settings. In contrast to other empirical findings though, our data reveals evidence of ex-ante moral hazard demonstrated by clear behavioural differences between those with self-funded supplemental health insurance and those for whom the workplace finances the additional insurance. We find no evidence that either form of insurance is related to improved self-assessed health.