Recent research suggests that an additional $1 of health aid would displace – or crowd out – nearly the same amount in a recipient government’s own health expenditure. Implementing a Sector Wide Approach (SWAp) may exacerbate crowding out because recipient governments should face fewer constraints when allocating health aid. This paper uses rigorous panel data methods to investigate this hypothesised effect of SWAps. We find that SWAps provide not an exacerbating but a potentially protective effect, reducing displacement of government health expenditure. This suggests some aid dollars are more fungible than others, and the mechanism for aid delivery makes a difference.