This paper examines the incidence of public health subsidies in Ghana using the GhanaLiving Standards Survey. Using a combination of (uniform) benefit incidence analysis and a discrete choice model, our results give a clear evidence of progressivity with consistent ordering:postnatal and prenatal services are the most progressive, followed by clinic visits, and then hospital visits. Children health care services are more progressive than adults'. Own price and income elasticities are higher for public health care than private health care and for adults than children. Poorer households are substantially more price responsive than wealthy ones, implying that fee increases for public health care will impact negatively on equity in health care.Simulations based on an estimated nested logit model show the importance of opportunity costs in healthcare decisions and suggest that reforms that focus only on out-pocket expenses will have a limited ability to extend public healthcare to all potential users.