In an ongoing population study in Malmö, Sweden, serum gamma-glutamyltransferase (GGT) was used both as an indicator of alcohol-related problems and as an aid in further investigation and treatment. The study group was consecutively selected from six middle-aged male birth-year cohorts of 8859 screening attenders (76.1% attendance), and comprised a random one-half (N = 317) of the individuals with screening GGT in the top decentile of the distribution. Alcohol consumption was the main cause of GGT elevation in 76% of the study group. Symptoms of alcoholism were frequent in these subjects but were rare in those with normal GGT levels. Marital and occupational measures were similar to those in the general population for subjects who were still in treatment but differed for subjects who had dropped out within 2 yr. Comparisons of medical and social data, the questionnaire results and other screening measures between the study group and the other subjects with screening GGT in the top decentile showed equivalence between the two groups and indicated that the study group was representative of the entire group with significantly elevated GGT values in the screening investigation.