AbstractThis study reports the continuous 8‐h monitoring of data on extremely low‐frequency magnetic fields (ELF‐MF) relating to 14 children and 35 teachers in 11 elementary schools in Northern Taiwan. It was anticipated that the subjects in two of these campuses would have elevated exposure to ELF‐MF as a result of their close proximity to high‐voltage (161 kilo‐Volt, kV) power lines. The results of our analysis reveal that in those schools with high‐voltage power lines running through the campuses, the mean ELF‐MF exposure level (0.38 ± 0.51 micro‐Tesla (µT), or 0.15, 0.25 and 0.44 µT at the respective 25th, 50th and 75th percentiles) was higher than the mean ELF‐MF exposure level for campuses situated far away from such high‐voltage power lines (0.14 ± 0.27 µT, or 0.04, 0.06 and 0.10 µT at the respective 25th, 50th and 75th percentiles). The multi‐level analytical technique, which takes individual measurements as the analytical unit, and which also takes into consideration the inter‐correlation between measurements from the same individual and/or campus, was also applied to the analysis of the data. We conclude that individual‐level and school‐level clustering of the measurements, both of which were discernible in this study, should be taken into consideration in any future analysis of data obtained from the continuous monitoring of exposure to ELF‐MF. Bioelectromagnetics 29:564–570, 2008. © 2008 Wiley‐Liss, Inc.