Lymphocyte cultures from the peripheral blood of 38 patients undergoing a cytostatic interval therapy with a regimen of methyl-CCNU (1-[2-chloroethyl-3-(4-methyl-cyclohexyl)]-1-nitrosourea), 5-fluorouracil, and vincristine (each 5-day course of therapy was followed by a therapy interval of 4 weeks) were supplied with 5-bromodeoxyuridine (BUDR) for the whole culture time to determine the sister chromatid labelling pattern. From a total of 92 individual blood samples sister chromatid exchange (SCE) studies were performed including analyses before the start of the therapy, and immediately and 4 weeks after each course of therapy. In addition, the frequency of first, second, and third metaphases in the 72-h cultures was estimated using the characteristic labelling patterns. A distinct increase of SCE frequency over the control level (i.e., lymphocyte cultures of patients before the start of therapy) was observed at all phases of therapy. It was clearly correlated with the number of courses of therapy up to course 7, later on the SCE rate remained more or less at the level reached. The influence of the composition of each drug regimen on the SCE rate was less pronounced than it was on the breakage rate. Moreover, although a clear correlation existed between the individual rates of breakage and SCE, the formation of the latter appeared to reflect a long-term effect of the therapy rather than did the formation of break aberrations. In addition, as the intercellular variability of the number of SCEs per cell was much higher than that of breaks, the interindividual variability (variation of the mean values for each patient) was small compared to the respective variability of breakage rates. The proportion of first, second, and third metaphases present in 72-h cultures evidently was influenced by single courses of therapy. The observed delay of proliferation was also reflected in different amounts of chromosome damage. Although the BUDR treatment enhanced the cytostatic effect of the therapy on the lymphocytes in culture rendering SCE analysis rather difficult in several cases, the other data of this study and in particular the experiences with the "long-term effect" make it imperative to include BUDR-labelling in further cytogenetic studies in subjects with exceptional exposure to chemicals. However, the SCE method can by no means, replace the classic cytogenetic analysis.