There has been recent controversy about whether aid directed specifically to health has caused recipient governments to reallocate their own funds to non-health areas. At the same time, general budget support (GBS) has been increasing. GBS allows governments to set their own priorities, but little is known about how these additional resources are subsequently used. This paper uses cross-country panel data to assess the impact of GBS programmes on health spending in low-income and middle-income countries, using dynamic panel techniques to estimate unbiased coefficients in the presence of serial correlation. We found no clear evidence that GBS had any impact, positive or negative, on government health spending derived from domestic sources. GBS also had no observed impact on total government health spending from all sources (external as well as domestic). In contrast, health-specific aid was associated with a decline in health expenditures from domestic sources, but there was not a full substitution effect. That is, despite this observed fungibility, health-specific aid still increases total government health spending from all sources. Finally, increases in total government expenditure led to substantial increases in domestic government health expenditures. Copyright © 2012 John Wiley & Sons, Ltd.