Summary The seasonality of presentation of 1019 skin melanomas in Oxford Region 1952Region -1975 Several recent studies (Fears et al., 1977;Houghton et al., 1978;Wigle, 1978;Swerdlow, 1979;Houghton et al., 1980;Houghton & Viola, 1981;MacKie & Aitchison, 1982) have suggested that cutaneous malignant melanoma incidence may have a short induction period after exposure to sun radiation. One finding which could be interpreted as supporting this hypothesis is the seasonal pattern of first diagnosis of melanoma found in one or both sexes in Sweden (Malec & Eklund, 1978), the Third National Cancer Survey (TNCS) areas of the United States (Scotto & Nam, 1980), Hawaii (Hinds et al., 1981) and Western Australia (Holman & Armstrong, 1981), and perhaps in Lane County, Oregon (Morton & Starr, 1979) -such seasonality of presentation could reflect real seasonality of incidence, which could most readily be explained by a short induction period from a seasonal aetiological insult such as ultraviolet radiation exposure. However, several alternative explanations of the seasonal presentation are possible and have not been fully explored in previous work: the disease might be noticed more rapidly in summer and hence present more often then; or the way in which medical care is organised or used might