CT is an important psychiatric treatment, particularly for severe or refractory major depression (1, 2). Data from states and localities suggest that after increasing during the 1940s and 1950s, use of ECT declined over the following decades (3-6). In analyses of national hospital survey data, Thompson et al. (7) found that use of ECT decreased 46% between 1975 and 1980 but that it did not significantly change between 1980 and 1986. The authors concluded that the decline in use of ECT in the United States ended in the 1980s. A subsequent analysis of Medicare claims (8) found that use of ECT began to rise again in the early 1990s among the nation's elderly and Medicare-eligible disabled.Trends in aggregate annual ECT use do not adequately reveal the underlying variation. The prevalence of major depression (the principal disorder treated with ECT) has been shown to be consistent across geographic areas (9). However, per capita utilization rates of ECT vary widely across the United States. A study of ECT use in 317 U.S. cities (10), based on survey data, found no use of ECT in one-third of cities. Among the cities where its use was reported, the rate in the city with the greatest use was more than fivefold higher than the rate in the city with the least use. Such patterns of use suggest a lack of access to the procedure in some areas and a lack of uniformity in its administration (11).The variation study found that one of the most influential predictors of the amount of ECT administered was an area's supply of psychiatrists. It is well-known that psychiatrists are unevenly distributed across the United States; for example, states vary more than eightfold in the number of psychiatrists per capita (12). Overall psychiatrist supply provides only a partial explanation, however, because a very small proportion of psychiatrists (less than 8% [13]) use ECT. In this report, we analyze the characteristics of psychiatrists who make use of ECT, in order to provide a fuller understanding of variation in its use. We also consider how changes in the psychiatric workforce may affect the availability of ECT.A literature search that used MEDLINE and other sources found little information about the characteristics of psychiatrists who provide ECT. A survey of a sample of members of the American Psychiatric Asso-