A successful in vivo application of the cytokinesis blocked micronucleus assay for the detection of aneuploidy induced by carbendazim (CARB) was carried out in the granuloma pouch assay. This was performed in two ways: (i) in vivo exposure of the skin fibroblasts to cytochalasin B (cytB) and CARB, by simultaneous injection of both substances into the pouch; (ii) in vivo exposure to CARB followed by in vitro culturing of the fibroblasts in the presence of cytB. Only the first assay was successful. Injection of cytB (with or without the test compound) into the pouch resulted in the induction of binucleate cells in vivo, up to a maximum of 5% at 1 mg cytB/pouch. After injection of CARB (0-50 or 0-10 mg/pouch) and cytB (1 mg) into the pouch, aneuploidy was determined in the isolated binucleate fibroblasts by fluorescence in situ hybridization with a general centromeric probe and combinations of chromosome-specific probes (19p + 19q, 4q + Yq). With all probes, the induction of chromosome loss and/or non-disjunction by CARB was very pronounced; at 10 mg CARB/pouch the total malsegregation frequency of chromosomes 4, 19 and Y was approximately 300/1000 binucleate cells. In an in vitro cytokinesis block assay with CARB (0-2.5 microg/ml) in primary skin fibroblasts the induced aneuploidy frequencies were as high as observed in the in vivo assay. The use of two probes for chromosome 19, which enabled the scoring of chromosome breaks in addition to aneuploidy, revealed no significant induction of chromosome breaks by CARB. The frequency of polyploid mononucleate and binucleate cells was decreased after CARB treatment, in both the in vivo and in vitro assays. However, in an additional in vitro assay without cytB a major induction of polyploidy from 2.5 microg/ml CARB and above was observed, showing that cytB may interfere with polyploidy induction.