The annual incidence rates of melanoma in Western Australia in 1975 and 1976 were estimated, by reference to hospital and pathology records. They were 4.4/ 100,000 in males and 6.2/100,000 in females for pre-invasive lesions and 18.6/100,000 and 18.8/100,000 for invasive lesions. Rates specific to particular body sites showed two distinct patterns of change with age - a progressive rise beginning about age 40 years (shown by Hutchinson's melanotic freckle and invasive melanoma of the head and neck) and an early rise beginning about age 20 years with a peak in middle-life and subsequent stabilization or decline (shown by superficial spreading melanoma, non-invasive and invasive melanoma of the lower limbs and, less typically, trunk and upper limbs). Incidence rates were highest in native-born Australians and higher in immigrants of British origin than in other immigrants. The rates in British immigrants were more than double those in the British Isles. The incidence was highest in residents of high social class areas, and in in-door rather than outdoor workers. Rates were also highest in the capital city, Perth, and the south-west corner of the State, rather than in the north where exposure of the population to ultra-violet light might be expected to be highest. In most of these respects the patterns for pre-invasive and invasive lesions were similar. Aspects of these data are inconsistent with the solar hypothesis of melanoma aetiology. It is suggested, however, that some of these inconsistencies may be explained if intermittent, intense exposure to the sun were more relevant to the aetiology of melanoma than continuous exposure.