An inaccurate evaluation of exposure is considered a possible cause for the inadequate conclusiveness of epidemiological research on adverse effects of extremely low frequency-magnetic fields (ELF-MF). The objective of this study is to provide an evaluation of current ELF-MF exposure in workers, the specific contribution of occupational exposure to overall 24-h exposure, and the representativeness of a job exposure matrix (JEM). ELF-MF exposure was monitored in 543 workers for 2 days using personal meters. Time-weighted average (TWA) levels at work, at home and outside the home were calculated. A JEM based on the 1988 International Standard Classification of Occupations (ISCO 88) was created. Median exposure at work, at home and outside the home were 0.14, 0.03 and 0.05 μT, respectively. Occupational exposure accounted for about 60% of 24-h exposure. In the JEM, about 50% of the classified occupations included significantly different individual TWAs. Occupational exposure to ELF-MF appeared low. Median exposure levels at home and outside were 20-28% of the occupational level, giving a minor contribution to overall day-to-day exposure. The frequent occurrence of workers with different TWA included under the same job title highlights the risk of misclassification in epidemiological studies on ELF-MF effects based on JEM.