All death records in Australia during the period 1964-1985 were analyzed to compare mortality from melanoma in immigrants and Australian-born individuals, and to investigate changes in risk in the immigrants according to their duration of stay and age at arrival. About 450,000 deaths were from cancer, and risks of melanoma were estimated by logistic regression relative to those of the Australian-born, with deaths from other cancers used as controls. Estimates were adjusted for age at death, time period, birth cohort, and state of registration of death in Australia. Region of birth was defined as New Zealand, other Oceania, England, Ireland/Scotland/Wales (including Northern Ireland and the Republic of Ireland), Central Europe, Eastern Europe, Southern Europe, Western Asia, or Eastern Asia, bearing in mind that many migrants born in Asia were of European descent. Overall, migrants from outside Oceania were at lower risk than the Australian-born, and the lowest risks in males were in Southern Europeans and Eastern Asians, reflecting the protective effect of a darker complexion. Risk of melanoma was related both to duration of stay in Australia and to age at arrival; although their relative importance cannot be measured, the patterns of change suggested that childhood migration may be more important in determining risk than number of years in Australia. The authors believe this study to be based on the largest data set ever used in migrant studies, and note that the previously found differences in melanoma risk between immigrants and Australian-born remained after adjustment for major temporal and geographic confounders. The results confirm the importance of the interaction between environmental and genetic risk factors in the etiology of melanoma.