After continuous labeling of proliferating (P) cells with 5-bromo-2'-deoxyuridine (BrdU) for 5 days, SCC VII tumor-bearing mice received various kinds of DNA-damaging treatments: gamma-ray irradiation, tirapazamine (TPZ, hypoxia-specific cytotoxin) administration, or cisplatin injection. From 0.5 to 72 hr after treatment, tumors were excised, minced, and trypsinized. Single tumor cell suspensions were incubated for 48 hr with a cytokinesis-blocker, cytochalasin-B. Then, the micronucleus (MN) frequency for BrdU-unlabeled cells, quiescent (Q) cells at treatment, was determined using immunofluorescence staining for BrdU. The MN frequency for total (P+Q) cells was obtained from tumors that were not pretreated with BrdU labeling. The sensitivity to each DNA-damaging treatment was evaluated in terms of the frequency of induced micronuclei in binuclear tumor cells (MN frequency). Treatment with gamma-rays or cisplatin resulted in a larger MN frequency in total cells than in Q cells. In contrast, TPZ treatment produced a smaller MN frequency in total cells than in Q cells. Regardless of the treatment used, Q cells showed greater repair capacities than total cells. However, TPZ caused much smaller repair capacity in both total and Q cells, compared with gamma-rays or cisplatin. Gamma-rays and cisplatin produced similar repair patterns. Differences in sensitivity between total and Q cells and repair patterns of the two cell populations were thought to depend on differences between the two cell populations in the toxicity of the DNA-damaging treatment and distribution pattern of the anticancer agent.