Several groups of investigators have reported that the extent of repair replication following treatment of mammalian cells with dimethyl sulfate is increased by 3-aminobenzamide (3AB) over that occurring in its absence. Two plausable explanations for this phenomenon were tested. The first is that the number of nucleotides inserted per repair site, that is the repair patch size, is increased and the second is that the rate of repair is increased. Human T98G cells were treated with dimethyl sulfate and allowed to repair their DNA in the presence or absence of 3AB. It was found that the presence of 3AB did not increase the repair patch size nor the rate of removal of methylation products from the DNA, which is assumed to equal the rate of repair. The results of further experiments suggested that the 3AB effect can be explained by changes in nucleotide precursor pools.