A questionnaire was mailed to a random sample of 771 undergraduate women at Virginia Tech. The research instrument was designed to accurately estimate the frequency of bulimia among university women. Bulimia was operationalized in specific behavioral terms that extend the DSM-III definition. The response rate for the mail survey was 87%, with 669 surveys used in the descriptive data analysis. One hundred and twelve subjects, 17% of the total sample, were classified as bulimic. The bulimic subjects were matched on height, weight, and age with 112 women without an eating disorder. This subset of 224 subjects was analyzed in order to assess the degree to which the respondent's family system, use of coping strategies, and weight preoccupation are predictive of bulimia. Multiple regression analysis indicated that 28% of the variance in bulimia can be explained by these three variables. The survey findings and implications for future research are discussed. / Ph. D.