The fate of cyclophosphamide (50180) (CP) induced lesions causing sister chromatid exchanges (SCEs) in mouse bone marrow and spleen lymphocytes under in-vivo and in-vivo/in-vitro conditions were compared. Measurements of SCEs were taken from 6 to 48 hours following treatment of male CD1-mice with CP injected intraperitoneally at 40mg/kg body weight. For in-vivo assay, mice were treated with 5-bromodeoxyuridine and colchicine prior to sacrifice, and bone marrow and spleen cells were removed and evaluated. For in-vivo/in-vitro assay, cells were removed and cultured with 5-bromodeoxyuridine and colchicine for specific times before analysis. SCEs increased rapidly in-vivo following treatment in both bone marrow and spleen cells. The number of SCEs peaked by about 12 hours after exposure and remained relatively unchanged until 24 hours. At this time, they began to decrease and reached a minimum at 48 hours which was still greater than controls. The importance of length of time for complete metabolic activation, which would influence pharmacokinetics, was indicated by the time dependent increase in SCEs. The proliferative nature of bone marrow and spleen cell populations sampled and resultant dilution of DNA damage over time were primary factors in lack of accumulation of SCEs over longer times. In-vivo/in-vitro analysis showed an increase in SCEs up to 6 hours which was more marked in spleen cells, followed by a decrease in bone marrow to 18 hours. Spleen cell SCEs decreased more slowly and remained elevated at 48 hours, when bone marrow numbers were similar to controls. Comparison of in- vivo and in-vivo/in-vitro tests indicated a significant difference, which varied with tissue. This was attributed to culturing and/or cell cycling effects since in-vivo/in-vitro cells had an opportunity to undergo two extra cycles. The authors conclude that in-vivo/in- vitro test results are similar to in-vivo results for cyclophosphamide. However, in-vivo/in-vitro tests may not indicate the true extent of in-vivo damage. Factors such as pharmacokinetic properties, sampling time, and tissue analyzed must be considered in cytogenic monitoring.