Melanoma risk differs by sun-sensitive phenotype, of which self-reported skin type (tendency to burn, inability to tan) is an indicator. If self-reports of skin type are influenced by the amount of sun exposure subjects have had, the two principal determinants of risk are linked, and stratifying by skin type would bias the estimated effect of sun exposure. Our objective was to determine whether teenagers changed their self-reports of skin type after being sunburnt. A random sample (n = 364) of 14-15 year old schoolchildren of northern European ancestry self-assessed and self-reported their skin type before and after their summer holidays. Their responses had high correlation coefficients (girls 0.71, boys 0.54) for repeatability at 4 months, but subjects who sunburnt less (more) frequently than usual that summer revised their skin type assessment to be less (more) sun-sensitive afterwards. We conclude that these 14-15 year olds were influenced by a recent experience of sunburn when reporting their skin type. A more objective measure of phenotype is needed.