A comparison of DNA polymerase III core enzyme (McHenry, C. S., and Crow, W. (1979) J. Biol. Chem. 254, 1748-1753) prepared from wild type Escherichia coli and a strain harboring the mutator gene, mutD5 (Degnen, G. E., and Cox, E. C. (1974) J. Bacteriol. 17, 477-487) has revealed several differences in their properties. Among these are alterations in the heat stability, divalent cation requirement, pH optimum, 3'----5'-single strand exonuclease activity, and DNA-dependent conversion of a deoxynucleoside triphosphate to its corresponding monophosphate ("turnover"). The decrease in the 3'-single strand exonuclease and turnover indicate a defect in the editing function of the mutD strain, which is at least in part responsible for the high spontaneous mutation rate in mutD. Transformation of mutD by a hybrid plasmid, pRD3, constructed from an EcoRI restriction fragment of E. coli and pBR322, cures mutD of its abnormally high mutation rate, and simultaneously restores its 3'-exonuclease activity. These observations are consistent with the notion that the mutD gene product is a subunit of DNA polymerase III, and it either contains the catalytic site for the 3'-exonuclease or modulates its activity. From a consideration of the known molecular weights of the subunits in DNA polymerase III core (McHenry C. S., and Crow, W. (1979) J. Biol. Chem. 254, 1748-1753) the molecular weights of the two proteins translated in maxicells transformed with pRD3, and from a comparison of our results with those obtained with the mutator dnaQ (Horiuchi, T., Maki, H., Maruyama, M., and Sekiguchi, M. (1981) Proc. Natl. Acad. Sci. U. S. A. 78, 3770-3774) and the work of Cox and Horner (Cox, E. C., and Horner, D. L. (1983) Proc. Natl. Acad. Sci. U. S. A. 80, 2295-2299) as well as Echols et al. (Echols, H., Lu, C., and Burgers, P. M. J. (1983) Proc. Natl. Acad. Sci. U. S. A. 80, 2189-2192) we tentatively assign the mutD gene product to the epsilon subunit of DNA polymerase III.