Abstract Micronucleus (MN) induction as a genotoxic effect of extremely-low-frequency electromagnetic fields (ELF-EMF, 50 Hz, 1 mT) was studied in human amniotic fluid cells (AFC) after continuous exposure to magnetic fields (MF), oriented horizontally and vertically with respect to the surface of the culture medium, at different time points. To compare the effectiveness of different exposure systems, a Helmholtz-coil system and a so-called Merritt-coil system was used. A statistically significant increase in MN frequency could be detected in exposed cells compared to controls after 72 h continuous exposure to MF applied vertically in the Merritt-coil system, while no effect was found after exposure in the Helmholtz-coil system. Furthermore, a significant increase in MN induction occurred after 24, 48 and 72 h exposure to MF applied horizontally in the Helmholtz-coil system in comparison to controls, whereas horizontally MF generated in the Merritt-coil system induced no genotoxic effects. To exclude suppression of indirect EMF-induced DNA-lesions, we studied MN formation in the presence of N-Acetyl- p-aminophenol (APAP, Paracetamol®), which is an inhibitor of DNA-repair mechanisms. We found a dose-dependent increase of MN formation in APAP-treated AFC cells, but no significant further increase in MN frequency after additional MF exposure. Therefore we conclude, that EMF-induced MN formation is not caused by directly or indirectly induced clastogenic mechanisms. The obtained results show that the orientation of MF with respect to the cell culture dish and the physical condition of the exposure system is of major importance for the induction of micronuclei in certain cell types. Therefore, the reason for inconsistent results published in the literature may be caused by the variability of exposure systems, the exposure conditions and the cell types used.