Privatization in Estonia has produced varied ownership configurations. This enables hypotheses on the productivity effects of different ownership forms to be tested. Findings are based on fixed‐effects production function models and are estimated using a large, random sample of firms. Depending on the particular specification (and relative to state ownership), (i) private ownership is 13–22% more efficient; and (ii) all types of private ownership are more productive, though managerial ownership has the biggest effects (21–32%) and ownership by domestic outsiders has the smallest impact (0–15%). The joint hypothesis that privatization coefficients are equal is rejected. Findings are robust with respect to choice of technology and the use of instrumental variable estimates. These results provide only partial support for the standard theory of privatization, but stronger support for theorists who argue that some forms of insider ownership may constitute preferable forms of corporate governance in some circumstances.