To determine the rate of appearance of nevi in a cohort of children between ages 6 and 12 years. The number of nevi has been established as the strongest known risk factor for melanoma, but whether the rate at which nevi appear during childhood varies by age is not well understood. The study involved analysis of nevus development over time in a cohort of 640 Western Australian school children who formed the control group in the Kidskin intervention trial. Children were assessed at ages 6, 10 and 12 years, with nevi on the back counted from photographs. The changes in both number and density of nevi between ages 6 and 12 years were linear with respect to age. The number and density of nevi increased at a greater rate for boys compared with girls; for children with blue, hazel and green eyes compared with those with brown eyes; for blonde versus dark haired children and for children with heavy freckling compared with those with no freckles. In conclusion, further research is needed to determine whether the linear increase in nevi continues into adolescence. If so, it suggests that skin cancer prevention campaigns should target older children and adolescents as well as younger children. Longitudinal studies have important advantages over cross-sectional studies in the description of nevus development with age, and should be the preferred study design.