A micronucleus assay employing fluorescence in situ hybridization (FISH) with a centromeric probe was used on specimens of exfoliated buccal and nasal cells collected from mortuary science students exposed to embalming fluid containing formaldehyde. FISH labeling allowed micronuclei (MN) containing a whole chromosome (centromere-positive, MN+) to be differentiated from those containing only chromosomal fragments (centromere-negative, MN-). Each student was sampled before and after the 90 day embalming class. We determined if an increase in MN frequency could be attributed to formaldehyde exposure and was specific to either MN+ or MN-. In buccal cells, total MN frequency was significantly increased from 0.6/1000 to 2/1000 (p = 0.007) following the course, whereas in nasal cells it was not (2 and 2.5/1000, respectively, p = 0.2). Cells with multiple MN were present only in samples taken after exposure to embalming fluid. Although the baseline frequency was higher for MN+ in both buccal (0.4/1000 for MN+ and 0.1/1000 for MN-) and nasal cells (1.2/1000 for MN+ and 0.5/1000 for MN-), the increase in MN frequency was greater for MN-, (9-fold, p = 0.005 for buccal cells; 2-fold, p = 0.03 for nasal cells) than for MN+ (> 2-fold, p = 0.08 for buccal cells; no change, p = 0.31 for nasal cells) in both tissues. Thus, the primary mechanism of micronucleus formation appeared to be chromosome breakage. This finding is consistent with known clastogenic properties of formaldehyde, the component of embalming fluid most likely responsible for micronucleus induction.