The influence of sampling time on the frequencies of micronuclei, centromere-positive micronuclei and chromosome nondisjunction was investigated in binucleated lymphocytes following treatment with a known clastogen (mitomycin C) or an aneuploidy-inducing agent (vincristine sulfate). Cytochalasin B (6 micrograms/ml) was added 44 h after mitogen stimulation and cultures were harvested 12, 28, 36 and 48 h thereafter. Micronucleated cells and micronuclei were significantly induced by the two treatments at all sampling times. Furthermore, in situ hybridization with an 'all centromeres' probe showed that vincristine-induced micronuclei were prevalently centromere-positive whereas in mitomycin C-treated cultures only a minor fraction of induced micronuclei contained the hybridization signals. Chromosome nondisjunction rates, as measured by in situ hybridization with chromosome 7- and 11-specific alphoid probes, significantly increased following vincristine treatment. Chromosome nondisjunction and total micronucleus frequencies were found to increase with time both in controls and in mutagen-treated cultures, whereas the percentage of centromere-positive micronuclei in the different treatments was not influenced by the sampling time. Our data suggest that even in the presence of 6 micrograms/ml cytochalasin B, the abnormal segregation of binucleated cells may contribute to the baseline level of micronuclei and influence the results obtained. The introduction of a short cytochalasin B treatment (between 12 and 28 h) in the cytokinesis-blocked micronucleus assay may avoid the cytochalasin B effect on micronucleus frequencies.