Objective: To examine associations between early-life sun exposure and risk of invasive cutaneous melanoma diagnosed between ages 18 and 39 years. Methods: Data were analysed from 606 cases and 481 controls from the Australian Melanoma Family Study, a population-based, case-control-family study. Self- and parent-reported sun exposure was collected by interview. Odds ratios (OR) were estimated using unconditional logistic regression, adjusted for potential confounders. Results: Self-reported childhood total sun exposure was not associated with melanoma overall, but was positively associated with melanoma diagnosed at 18-29 years of age (OR for highest vs. lowest quartile: 3.21, 95% confidence intervals (CI) 1.38-7.44; P trend 0.02; P interaction by age group 0.09). Analyses restricted to participants whose self-reported sun exposure was concordant with that recalled by their parents gave an OR for the highest versus lowest tertile of childhood total sun exposure of 2.28 (95% CI 1.03-5.04; P trend 0.05), and for any versus no severe childhood sunburn of 2.36 (95% CI 1.05-5.31). The association of self-reported severe sunburn with melanoma was evident only in people who tended to tan rather than burn and in people who had few nevi. Conclusion: The association of early-life sun exposure with early-onset melanoma is influenced by host factors.