Although the PIG-A gene mutation frequency (MF) is considered a good proxy to evaluate the somatic MF in animals, evidence remains scarce in humans. In this study, a granulocyte PIG-A-mutant assay was evaluated in patients undergoing radiation therapy (RT) for breast cancer. Breast cancer patients undergoing adjuvant RT were prospectively enrolled. RT involved the whole breast, with (WBNRT) or without (WBRT) nodal area irradiation. Blood samples were obtained from participants before (T0) RT, and T1, T2, and T3 samples were collected 3 weeks after the initiation of RT, at the end of RT, and at least 10 weeks after RT discontinuation, respectively. The MF was assessed using a flow cytometry protocol identifying PIG-A-mutant granulocytes. Cytokinesis-blocked micronucleated lymphocyte (CBML) frequencies were also evaluated. Thirty patients were included, and five of them had received chemotherapy prior to RT. The mean (±SD) PIG-A MFs were 7.7 (±12.1) per million at T0, 5.2 (±8.6) at T1, 6.4 (±8.0) at T2 and 3.8 (±36.0) at T3. No statistically significant increases were observed between the PIG-A MF at T0 and the MFs at other times. RT significantly increased the CBML frequencies: 7.9 ‰ (±3.1‰) versus 33.6‰ (±17.2‰) (p < .0001). By multivariate analysis, the CBML frequency was correlated with age at RT initiation (p = .043) and irradiation volume at RT discontinuation (p = .0001) but not with chemotherapy. RT for breast cancer therapy failed to induce an increase in the PIG-A MF. The PIG-A assay in humans needs further evaluation, in various genotoxic exposures and including various circulating human cells.