We propose a detailed analysis of heterogeneity in households’ responses to water conservation programs (price increase, voluntary and mandatory conservation) during periods of water shortage. Using a unique dataset covering water consumption of all residential users in Los Angeles (California) during the drought (1988-1992), we show that households generally were responsive to the conservation measures but that the magnitude of households’ responses varies depending on the instrument and on households’ characteristics, in particular the size of their lot. Price elasticity is estimated between –0.29 and –0.47 in the high season (June-October), and between 0 and –0.19 in the low season (November-May). Results suggest that the voluntary conservation program [resp. mandatory conservation program] induced a reduction in water use which varies from 1 to 13% [resp. 21% to 29%] depending on the season and the size of the lot. The achieved reduction in consumption is however very similar across households. These data also allow us to compare the effectiveness of price and nonprice policies in terms of water savings. Finally, welfare calculations suggest that households with the smallest lot sizes (and lowest income) suffered the greatest loss during the implementation of the water conservation programs.