To identify loci affecting disease susceptibility, linkage and association analyses are performed by scoring previously characterized sequence variation, such as microsatellites and single-nucleotide polymorphisms. However, such analyses can be expensive and time-consuming. Moreover, examination of particular candidate gene regions for potential new linkages and associations may be limited by the infrequent occurrence of known markers. The typing by currently available methods of a sufficiently large number of markers to achieve a robust analysis is often beyond the reach of the typical research laboratory. Various techniques that score either unknown or known sequence variation have been developed, yet none of these strategies is equally well suited for mutation detection and the scoring of genotypes. To perform both tasks, a technique should be both sensitive enough to detect all mutation types and sufficiently quantitative that the translation of the data into allele-sharing status is feasible. We investigated the possibility of replacing conventional genotyping with a mutation-detection approach. Genome mismatch scanning is a hybridization-based mutation-detection technique employing the Escherichia coli mismatch-detection enzymes MutHLS. It was originally developed to enrich for identical-by-descent regions between two whole genomic DNA samples. To accommodate polymerase chain reaction products of candidate regions, we have conducted a comprehensive biochemical optimization of the technique for target sizes ranging from 260 to 1,250 base pairs. Our modifications, which we now collectively designate polymerase chain reaction–candidate region mismatch scanning, have simplified the assay, rendered it quantitative and demonstrated its potential for cost-effective, high-throughput genotyping. Strategies for exploiting the method in the study of candidate regions or in genome-wide studies will be discussed.