Abstract How effective are capital account restrictions? We provide new answers based on a novel panel data set of capital controls, disaggregated by asset class and by inflows/outflows, covering 74 countries during 1995–2005. We find the estimated effects of capital controls to vary markedly across the types of capital controls, both by asset categories, by the direction of flows, and across countries' income levels. In particular, both debt and equity controls can substantially reduce outflows, with little effect on capital inflows, but only high-income countries appear able to effectively impose debt (outflow) controls. The results imply that capital controls can affect both the volume and the composition of capital flows.