Serum antibodies against human coronavirus OC43 in different age groups were measured by complement fixation (CF), haemagglutination inhibition (HI), radial diffusion haemolysis-in-gel (HIG), and solid-phase radioimmunoassay (RIA) methods. Antigen grown in suckling mouse brain was used in all tests. Results obtained by the CF and HIG tests, and the RIA, were in good agreement with regard to the presence or absence of antibodies. Similar results were also obtained with the HI test if nonspecific haemagglutination inhibitors were first removed by treatment with phospholipase C and only titers of 1:20 or greater were considered positive. Children 6--23 months of age (n = 45) were without measurable coronavirus antibodies in all four assays. A rapid increase in the prevalence of antibodies then occurred in subsequent age groups, and practically all persons 6 years of age or older were found to have OC43 antibodies as measured by the HIG test or the RIA. The mean antibody levels determined by these two methods continued to increase, however, up to the age group of 10--14 years. This increase in antibody levels after the initial antibody incidence plateau may be due to boosting effects caused by related coronavirus strains, since OC43 antigens are known to cross-react with antibodies induced by other human coronaviruses. Taken together, these data suggest that OC43 virus, or an antigenically related coronavirus strain, is very common in Finland.