Melanoma risk among subjects from Germany, France and Belgium who had lived for 1 year or more in sunny climates was examined in a one-to-one unmatched case-control study conducted among white subjects 20 years old or more. A total of 412 consecutive patients with melanoma diagnosed from 1 January 1991 onwards, were derived from hospital registers; 445 controls were randomly chosen in the same municipality as the cases. After adjustment for host characteristics, melanoma risk associated with residence in a sunny area was 2.7 (95% CI: 1.4-5.2), increasing to 4.7 (95% CI: 1.4-13.5) if subjects sought a suntan when residing in sunny climates, and to 4.3 (95% CI: 1.7-11.1) if subjects arrived before the age of 10 years in the sunny area. Residence in sunny areas and recreational sun exposure seemed to combine their effects on melanoma risk. Increase in melanoma risk conveyed by deliberate sun exposure during adulthood was highest among subjects who had lived in sunny areas as a child or adolescent and lowest among subjects who had never resided in sunny areas. Our results support conclusions from migrant studies that indicated that childhood is a critical period of either vulnerability to solar radiation or more frequent exposures to melanoma risk factors. They also suggest that moderate sun exposure of an adult who was heavily sun exposed in childhood is associated with a higher melanoma risk than that of high sun exposure of an adult who was sun protected in childhood.