Bleomycin (BLM) induced chromosomal damage in G2 phase and its repair kinetics in normal human lymphocytes were studied following different treatment schedules. As a first step, a dose-response curve was obtained (concentrations of 5-50 micrograms/ml). For repair kinetics studies, blood samples were treated with BLM at a concentration of 20 micrograms/ml. Continuous treatment produced equal numbers of breaks per cell (br/c) when the cells were treated 3, 4 or 5 h before fixation. If the treatment time was extended to 6 h, the level of br/c was increased 2-fold (p < 0.001) as a result of an increased number of cells with more than 3 br/c. The curves obtained after pulse treatment showed maximal chromosome damage at time 3 (45 min BLM treatment, followed by 2 h repair in drug free medium). When the time after treatment was extended to 4 h (treatment time 5), a 50% reduction in chromosome damage was measured. It was found out that at treatment points 3, 4 and 5 the differences in breaks per cell at the different schedules applied were statistically highly significant. If caffeine (CAF) was added, the continuous treatment, BLM+CAF, induced a statistically significant increase in the frequency of br/c at every treatment point, but the shape of the curve illustrating the kinetics of chromosomal damage remained unchanged. Moreover, the addition of CAF at continuous BLM treatment brings the level of br/c close to that measured at the pulse BLM treatment except for treatment time 3. When applied in a combination with BLM, CAF considerably modified the kinetics of chromosome damage for a pulse (BLM alone) treatment. The possible reasons for the changes in the level of br/c as well as a tentative scheme for assessment of chromosome damage repair capacity after BLM treatment are discussed.