The rodent Pig-a assay has been used extensively as a potential regulatory assay for evaluating the in vivo mutagenicity of test substances. Although the assay can be conducted in different mammalian species, there have been only a few reports describing its use in humans, and rarely in genotoxicant-exposed human populations. In this study, PIG-A mutation frequencies (MFs) were evaluated in 36 azathioprine (AZA; human carcinogen)-treated inflammatory bowel disease (IBD) patients and 36 healthy volunteers. IBD patients exhibited a slight but statistically higher MF (6.10 ± 4.44 × 10-6 ) than healthy volunteers (4.97 ± 2.74 × 10-6 ) (P = 0.0489). The estimated relative risk for the exposed patients was 1.22 which indicated that AZA is a risk factor for inducing PIG-A mutation. However, the PIG-A MF showed no associations with AZA treatment duration or total AZA exposure. In addition, we performed the cytokinesis-block micronucleus test on the same samples. The frequencies of micronuclei (MN) and nuclear buds (NBUD) in IBD patients (MN: 4.70 ± 2.86‰; NBUD: 1.89 ± 0.95‰) were significantly higher than in healthy volunteers (MN: 1.47 ± 0.77‰, P < 0.001; NBUD: 0.90 ± 0.58‰, P = 0.004). MN frequency also had significant correlations with AZA treatment duration (P = 0.011) and total AZA exposure (P = 0.018). Our findings indicate that AZA-treated IBD patients have only a marginally significant increase in PIG-A MF; in contrast, a much stronger AZA-associated increase in genotoxicity was detected with the lymphocyte MN assay. Environ. Mol. Mutagen. 2019. © 2019 Wiley Periodicals, Inc.