The first detailed epidemiological and histological studies of cutaneous malignant melanoma in Western Australia have been undertaken recently. High and, apparently, increasing annual incidence rates have been confirmed (1975/76: 23.0/100 000 in males, 25.0/100 000 in females. 1980/81: 28.9/100 000 in males, 31.5/100 000 in females--rates for pre-invasive and invasive lesions combined). The results support a causal relationship of sunlight exposure with cutaneous melanoma in general, while suggesting that melanoma of Hutchinson's melanotic freckle type is related to continuous sun exposure whereas intermittent exposure is more important in the etiology of melanoma of superficial spreading type. The proposed etiological heterogeneity of melanoma and the dual origin theory of Mishima have been embodied in a theory of the etiology and histogenesis of melanoma which proposes that nodular melanoma is a common end result of the other types of melanoma rather than a distinct histogenetic entity. Some components of this theory have been supported by results of the 1980/81 West Australian Lions Melanoma Research Project. The overriding importance of tumour thickness as the most accurate histologic index of prognosis yet available has been emphasized by correlation with survival rates in Western Australia and the Oxford Region. Tumour thickness has also been shown to be the most reproducible of histological features, while others were subject to considerable interobserver variation. The better prognosis for melanoma patients in Western Australia compared with low incidence regions, when correlated with tumour thickness, indicated that melanoma is diagnosed earlier in areas of high incidence due to greater awareness of the risk.(ABSTRACT TRUNCATED AT 250 WORDS)