Relationships between alcohol consumption and suicide rates were examined through an ecological study. Data on total apparent per capita alcohol consumption, as well as beer, wine, and spirits consumption, from 1977 to 1988 in each of the fifty U.S. states and District of Columbia were compared with suicide rates. It was found that total alcohol consumption was weakly but significantly related to suicide rate by univariate analysis (r = 0.16, p < 0.01). Pearson correlation coefficients were 0.30, 0.12, and -0.01 for correlation of suicide rate with beer, wine, and spirits consumption, respectively. Over the twelve years, the magnitude of associations between total alcohol consumption and suicide varied with correlation coefficients of 0.29, 0.23, 0.09, and 0.11 for four three-year periods, 1977-79, 1980-82, 1983-85, and 1986-88, respectively. The association between beer consumption and suicide persisted over the whole period with correlation coefficients of 0.35, 0.31, 0.26, and 0.30, while that between wine and suicide did not. Associations were also examined by quartile of each alcohol's consumption. Significant inverse association was found for beer as well as total consumptions at the lowest quartile. However, at the highest quartile, beer and spirits consumption showed positive association with suicide. These data suggest a "U-shaped" curve relating suicide rate to alcohol consumption. By using 1980 census data with adjustment for confounding effects of socioeconomic factors, total alcohol consumption was found to be positively associated with suicide especially at high consumption level. At low level consumption, beer showed an inverse association with suicide after controlling for confounding effects. © 1996, Baywood Publishing Co., Inc.