DNA single-strand breaks (about 200–300 per genome) were transiently detected during the first hour when HeLa cells were incubated for up to 24 h with 100 μM methotrexate. There was an expected increase in ADPribosyltransferase activity, which reached a maximum 2–3-fold stimulation at 3 h but which was still greater than in control cells after 24 h. When hypoxanthine (25 μM) was present in the incubations together with the methotrexate the transferase was no longer activated, although basal, control levels of activity were still present. DNA strand breaks were reduced in number but were still just detectable under these conditions. Cellular NAD + levels were mostly unaffected by the various drug treatments, except for a small transient decrease after 1 h, possibly as a result of the transferase activation. Methotrexate did not cause an increase in the rate of ADPribose degradation. Degradation of ADPribose residues labelled in a preincubation period in permeabilized cells was more extensive at pH 6.0 than at pH 8.0, but was slow relative to other published reports. The maximum rate observed at pH 6.0 was a 50% loss of acid-insoluble radioactivity in 30 min at 26°C. At pH 8.0 the loss did not exceed 30–35% even after 90 min incubation. The activation of the transferase is reflected in a general increase in protein ADPribosylation detected by autoradiography of 32P-labelled proteins in 6.25–18.25%T gradient acrylamide gels. There were three major acceptors with molecular masses of 17, 100 and over 100 kDa, which could be respectively a histone, a transferase-derived peptide fragment and the transferase itself. When ADPribosyltransferase was inhibited with 3-aminobenzamide DNA single-strand breaks were no longer detected. However, this had no observably significant effect on the kinetics of loss of cell viability (from Trypan blue uptake), cell number or colony-forming ability. Similar results are observed in most cases when the activation of the transferase, resulting from the incubation of cells with methotrexate, is inhibited by hypoxanthine. We conclude from such observations that the enhanced protein ADPribosylation seen in the cells exposed to methotrexate is a direct consequence of drug-exposure, but does not have any significant influence over the course of events leading ultimately to cell death.