Why does a large proportion of India’s urban poor choose to pay for private health services when public services are essentially free of charge? Drawing from previous research, this study takes the underprovision of the medication component in public health packages as a starting point. Field evidence shows that patients in public health facilities often have to get external medication, whereas private hospitals offer a menu of pre-specified care packages. Thilo Klein contributes an answer as to why the poor choose to go private by investigating their risk-aversion and demand for insurance in the choice of health facilities. In discrete choice experiments on maternity care in the slums of Hyderabad, the author finds that the lowest income group attaches significantly higher importance to “full medication” maternity care packages. In line with Expected Utility theory, the study further finds empirical evidence that this insurance demand of the poor is partially explained by their risk-aversion. The findings suggest that an effective pro-poor policy should provide insurance cover for medication-related expenses. They have significant implications for the provision of public health care to the poor and are important for policy-makers and health management alike.