We study the effects of a change in the way patient reimbursements are calculated on the prices of pharmaceuticals using quasi-experimental data for Denmark which switched from external (where reimbursements are based on prices of similar products in foreign countries) to internal reference pricing (where they are based on the cheapest domestic substitute). We analyze three therapeutic classes with different treatment durations and show that the reform led to substantial price decreases for our lifelong treatment and to less substantial price reductions for our medium duration treatment while we do not find significant effects on our acute treatment. Moreover, the reform did only affect generics and did not impact original products or parallel imports.