Little is known about the real everyday exposure of children in Europe to extremely low-frequency magnetic fields (ELF-MFs). The aims of this study are to (i) assess personal ELF-MF exposure in children; (ii) to identify factors determining personal and bedroom ELF-MF exposure measurements in children; (iii) to evaluate the reproducibility of exposure summary measures; and (iv) to compare personal with bedroom measurements. In Switzerland and Italy, 172 children aged between 5 and 13 years were equipped with ELF-MF measurement devices (EMDEX II, measuring 40–800 Hz) during 24–72 h twice, in the warm and the cold season. In addition, 24-h measurements were taken in the bedroom of children. In our study, sample geometric mean ELF-MF exposure was 0.04 μT for personal and 0.05 μT for bedroom measurements. Living within 100m of a highest voltage power line increased geometric mean personal exposure by a factor of 3.3, and bedroom measurements by a factor 6.8 compared to a control group. Repeated measurements within the same subject showed high reproducibility for the geometric mean (Spearman's correlation 0.78 for personal and 0.86 for bedroom measurements) but less for the 95th and 99th percentile of the personal measurements (≤0.42). Spearman's correlation between bedroom and personal exposure was 0.86 for the geometric mean but considerably lower for the 95th and 99th percentiles (≤0.60). Most previous studies on ELF-MF childhood leukaemia used mean bedroom exposure. Our study demonstrates that geometric mean bedroom measurements is well correlated with personal geometric mean exposure, and has high temporal reproducibility.