The aim of the nested case-control study reported here was to test the hypothesis that exposure to electromagnetic fields of the type generated by high-voltage power lines increases the incidence of cancer in children aged 0–14 years. The study population comprised children who during at least one of the years 1960, 1970, 1980, 1985, 1987, or 1989 had lived in a census ward crossed by a high-voltage power line. The cases were diagnosed from 1965 to 1989 and were matched to controls by year of birth, sex, and municipality. Exposure to electric and magnetic fields was calculated by means of computer programs in which power line charactenstics and distance were taken into account. No association was found between exposure to time-weighted average exposure to magnetic fields and cancer at all sites, brain tumors, lymphoma, or leukemia. Cancer at other sites showed elevated odds ratios in the two highest exposure categories in some, but not all, measures of exposure. This study provides little support for an association between children's exposure to magnetic fields and cancer and no support for an association between leukemia and such exposure, but no firm conclusions can be drawn owing to the small numbers involved. Am J Epidemiol 1997; 145: 219-26.