The study examines the consequences of alternative public responses to drought shocks. It does so by drawing on household data from resettlement areas of rural Zimbabwe from 1992-93 to 1995-96 and the estimation of four behavioral relations: The determinants of crop income; the determinants of investment in livestock; the determinants of investment in agricultural capital stock; and the determinants of private transfers. This information is used to construct a series of simulations in which drought relief received in the aftermath of the 1994-95-the ex-post response-is made available to households in the form of agricultural capital stock and extension advice-an ex-ante action. Doing so is found to raise household welfare in nondrought years, but provides only limited protection in the aftermath of drought. © 2003 Elsevier Science Ltd. All rights reserved.