The cancer chemotherapeutic agent methotrexate inhibits the enzyme dihydrofolate reductase leading to a depletion of cellular reduced folates and inhibition of thymidylate synthase. A predicted consequence of this depletion of cellular thymidylate residues is that the nucleotide dUMP may be incorporated into DNA in place of dTMP. The subsequent cycle of uracil removal and reincorporation by an excision-repair pathway may cause DNA damage, a possible contributory factor leading ultimately to cell death. DNA. damage, in the form of single- and double-strand breaks, was -5 detected in HeLa cells treated with high dose (>10 M) methotrexate using the sensitive nucleoid sedimentation technique and alkaline filter elution. The maximum level of DNA damage, in the form of single-strand breaks, was detected after only 1 hour of drug incubation, but breaks were not detectable after this time, presumably because they had been repaired. DNA double-strand breaks were detectable from 18 hours onwards, along with a small level of single-strand breaks. The presence of hypoxanthine together with the irethotrexate to overcome the inhibition of de novo purine biosynthesis led to a similar occurrence of strand breaks, although a greater number of double-strand breaks was detected. The early appearance of single-strand breaks coincided with a substantial decrease in cellular dTTP levels and loss in colony-forming ability of the cells. No significant changes in cell viability, as measured by a dye-exclusion assay, were detected until at least 12 hours drug incubation. The greater level of double-strand breaks in cells grown with methotrexate and hypoxanthine coincided with a greater loss in colony-forming ability in these cells. 6. Changes in chromatin structure in cells treated with drug for 24 hours or longer were detected as a faster sedimentation of nucleoids compared to control cells. These changes were not due to the supercoiled state of the DNA but were related to the protein component of the nucleoid cage structure. 7. The enzyme ADP-ribosyltransferase was found to be stimulated in response to the appearance of single-strand breaks, reaching a maximum stimulation of 2 - 3 fold after 3 hours incubation with methotrexate. With hypoxanthine present no activation was detected, even though single-strand breaks were present. In neither case was there a change in the degradation rate of ADP-ribose protein conjugates nor a significant change in intracellular NAD+ levels. 8. In methotrexate-treated cells, the activation of the transferase led to an increase in the ADP-ribosylation of three specific proteins, probably by the addition of mono (ADP-ribose) residues. 9. The complete inhibition of ADP-ribosyltransferase activity by 3-aminobenzamide, or the prevention by hypoxanthine of its activation by methotrexate, did not have any significant effect on changes in cell number, cell viability or colony-forming ability in cells incubated with methotrexate. Thus ADP-ribosylation of proteins is not an effective component during the cytotoxic action of methotrexate.