We investigate the enzymatic basis for the inefficient extension of single base mismatches by DNA polymerase compared with the extension of correct base pairs. Inefficient mismatch extension could result from either a reduced binding of the enzyme to mispaired versus correctly paired DNA template-primer termini, or from a lowered intrinsic rate of extension of mispairs by a bound enzyme, or from a combination of both factors. Avian myeloblastosis reverse transcriptase is used to measure the affinities (equilibrium dissociation constants) for the four matched and twelve mismatched base pair configurations situated at a primer 3'-terminus. The binding affinities are analyzed by two different assays employing polyacrylamide gels. The first assay uses steady-state kinetics to measure the efficiency of elongating correct and incorrect base pairs and to evaluate the enzyme's dissociation constants for matched and mismatched termini. The estimated KD values obtained in the steady-state analysis fall within a range of approximately 0.1-20 nM. The efficiencies of extending two of the mispairs, G.G and C.C, are too low to allow a determination of KD by the kinetics method. The second assay uses equilibrium binding to measure the ratio of polymerase bound to matched compared with mismatched termini, KDright/KDwrong. The affinity ratios, including values for G.G and C.C mispairs, are in the range of about 0.4-4.2. While around 1 order of magnitude difference is observed in the relative binding affinities of the polymerase for matched and mismatched primer termini, the relative extension efficiencies vary over more than 5 orders of magnitude. Therefore, it appears that inefficient mismatch extension is caused primarily by a kinetic block inhibiting elongation from mispaired primer 3'-termini rather than to a difference in binding.