For this study of the year-by-year change in the incidence of carriage of or infection by penicillin-resistant, coagulase-positive or haemolytic staphylococci in man, 248 contributions to the medical literature have been analysed and grouped under three main headings: people in hospitals, people with clinical signs attending out-patients' wards or doctors' surgeries, and people not currently receiving medical care.This analysis shows that there has been a steady increase in resistant strains among persons of the first group from 1942 to 1955, followed by a slight decrease to 1959, possibly owing to more discriminating use of penicillin. A steady increase was found in the other two groups from 1942 to 1959, the rate of increase being more pronounced among the group with clinical signs.The authors draw attention to the possible influence of the ingestion of penicillin with food on the incidence of resistant strains; the occurrence of such strains in certain foods; and the need for standardized techniques for sampling, for testing for staphylococci and for determining resistance.