In order to characterize mutagenic translesion DNA synthesis in UVM-induced Escherichia coli, we have developed a high-resolution DNA replication system based on E. coli cell extracts and M13 genomic DNA templates bearing mutagenic lesions. The assay is based on the conversion of M13 viral single-stranded DNA (ssDNA) bearing a single site-specific DNA lesion to the double-stranded replicative form (RF) DNA, and permits one to quantitatively measure the efficiency of translesion synthesis. Our data indicate that DNA replication is most strongly inhibited by an abasic site, a classic SOS-dependent noninstructive lesion. In contrast, the efficiency of translesion synthesis across SOS-independent lesions such as O6-methylguanine and DNA uracil is around 90%, very close to the values obtained for control DNA templates. The efficiency of translesion synthesis across 3,N4-ethenocytosine and 1, N6-ethenoadenine is around 20%, a value that is similar to the in vivo efficiency deduced from the effect of the lesions on the survival of transfected M13 ssDNA. Neither DNA polymerase I nor polymerase II appears to be required for the observed translesion DNA synthesis because essentially similar results are obtained with extracts from polA- or polB-defective cells. The close parallels in the efficiency of translesion DNA synthesis in vitro and in vivo for the five site-specific lesions included in this study suggest that the assay may be suitable for modeling mutagenesis in an accessible in vitro environment.