Alcohol-related disorders belong to the spectrum of major non-infectious diseases in Western societies which can be prevented by means that have not yet been fully implemented. Total consecutive mortality in a population of 10353 middle-age males invited to take a part in a preventive medical population program in Malmö was followed up for 3.5-8.5 years (mean 4.5) after the time of invitation and analysed in relation to participation or non-participation and forensic or in-hospital autopsy. Entry characteristics in the 7935 males who attended the screening were compared in order to evaluate risk factor patterns for the major categories of premature death during the follow-up period. Even in the males participating in the screening, alcohol-related deaths (ARD) constituted a major mortality category, comprising 55 of 218 cases, whereas cancer comprised 61 and coronary heart disease (CHD) 50 of the premature deaths in this group. Both in the ARD and CHD categories of male premature mortality, significant and distinctly differential risk factor patterns were found; in CHD for smoking, cholesterol, serum triglycerides and systolic blood pressure, and in ARD for gamma-glutamyltransferase, questionnaire alcoholism screening test and, inversely, serum cholesterol and serum creatinine. In both groups of diseases, these risk factors could be combined into highly predictive multiple logistic risk factor functions. The discriminative power of this instrument was even higher in ARD than in CHD deaths. In consequence, these factors may be applied both as indicators of the ARD risk and as signals and instrument for directed preventive measures in analogy with previously well established and tested methods for the regulation of blood pressure, serum lipids, etc. in the conquest of the cardiovascular diseases.