To address the genotoxicity of in vivo methyl bromide (CAS 74-83-9) exposure in humans, we collected blood and oropharyngeal cells as part of a cross-sectional morbidity study of methyl bromide-exposed fumigation workers and their referents. Micronuclei were measured in lymphocytes and oropharyngeal cells, and hypoxanthine-guanine phosphoribosyl transferase gene (hprt) mutations were measured in lymphocytes. A total of 32 workers and 28 referents provided specimens. Among current non-smokers, mean hprt variant frequencies (Vfs) were found to be elevated among workers compared to referents (geometric mean: workers=4.49x10(-6), referents=2.96x10-(6); two-sided p=0.22); this difference was more pronounced among workers with 4 h or more of recent methyl bromide exposure compared to referents (geometric mean: workers=6.56x10(-6), referents=2.96x10(-6); two-sided p=0.06). Mean oropharyngeal cell micronuclei were higher among workers compared to referents (mean: workers=2.00, referents=1.31; two-sided p=0.08); the results were similar when workers with 4 h or more of recent methyl bromide exposure were compared to referents (mean: workers=2.07, referents=1.31; two-sided p=0.13). No consistent differences between workers and referents were observed for frequencies of kinetochore-negative lymphocyte micronuclei, or kinetochore-positive lymphocyte micronuclei. The study was limited by a sample size sufficient only for detecting relatively large differences, absence of a reliable method to measure the intensity of workplace methyl bromide exposures, and relatively infrequent methyl bromide exposure (e.g., the median length of exposure to methyl bromide during the 2 weeks preceding the survey was 4 h). In conclusion, our findings provide some evidence that methyl bromide exposure may be associated with genotoxic effects in lymphocytes and oropharyngeal cells. Further study on the genotoxicity of methyl bromide exposure in humans is warranted.