In May, 1951, the Washington Public Opinion Laboratory conducted a study of 478 adults representative of the white population of the state of Washington. This sample estimated the frequency with which they, their friends, spouse, father, and mother drank alcoholic beverages. Using chi-square tests and T coefficients, the existence and degree of association between the drinking behavior of the sample and these other individuals was ascertained. Comparisons were made for the sample as a whole, and the sample classified by differences in sex, age, education, and rural-urban background. The findings showed that in general the drinking behavior of the sample conformed more closely to that of their contemporaries, friends or spouse, than to the behavior of members of the previous generation. Analysis in terms of the four characteristics, however, showed that this general relationship varied significantly in the different population segments.