The reliability and validity of sun exposure questions were examined in a group of 125 schoolchildren aged 14-15 years of northern European ancestry. We compared estimates of erythemally effective dose (EED) obtained from polysulphone badges worn on four consecutive weekend days in late spring to answers to several questions on habitual sun exposure. The polysulphone badge estimates accurately reflected the reported sun exposure of children and were highly correlated with responses to questions on habitual sun exposure obtained 12 months earlier. For the question "During weekends and school holidays, how much time do you usually spend in the sun each day", the Pearson correlation coefficients for the association with EED were 0.36 (girls) and 0.23 (boys). Adjusted for within-person variation in the EED measurements, the correlation coefficients were 0.65 (girls) and 0.43 (boys). For the question "Weekends and school holidays, where do you spend your time", the correlation coefficients were 0.18 (girls) and 0.32 (boys); the adjusted coefficients were 0.30 (girls) and 0.53 (boys). The findings suggest that "habitual' sun exposure in teenage children is a temporally stable behaviour that is reported with an acceptable degree of reliability and validity.