Several new methods of infertility treatment have been developed. There is therefore an increasing interest among physicians in gaining knowledge about the occurrence of infertility in order to estimate the need for treatment. This article examines epidemiological studies in industrialized countries 1970-1992 on the prevalences of infertility, involuntary infecundity, and the seeking of medical advice for infertility. The studies are compared on the basis of study design, definitions, methods of measurement and results. Comparison of the studies is difficult, as researchers use different concepts and there is a large variation in the delineation of the populations involved. The current prevalence of infertility among women in the fertile age group varies from 3.6 to 14.3%, the life-time prevalence varies from 12.5 to 32.6%, the prevalence of involuntary infecundity from 2.6 to 5.9%, and medical contact for infertility from 3.6 to 17%. It is probable that these large differences in prevalences are mainly due to differences in definitions and methods of measurement. We recommend that future studies be carried out in representative population groups, and that only couples who have tried to have children should be included in the estimated prevalences.