When a disease is controlled by two or more mendelian loci acting epistatically, it can be modeled in a linkage analysis as a single-locus mendelian disease with reduced penetrance. Previous work has demonstrated the reliability of such approximation for nuclear families, but not for extended pedigrees. We simulated extended pedigrees under two-locus models, in which one of the two disease loci was linked to a marker, and analyzed them both under the correct two-locus models and under single-locus approximations. The single-locus approximations provided results that were very close to the correct two-locus results. This held true, whether we ascertained pedigrees based on the presence of at least one affected individual, or based on the presence of at least five affected individuals. While a simulation study cannot guarantee that extrapolation of the results to models other than those examined is justified, our findings strongly suggest that single-locus linkage analysis can be reliably used in analyzing two-locus disorders in extended pedigrees. We also found striking confirmation of the importance of performing linkage analyses under both dominant and recessive models when the mode of inheritance is unknown, for extended pedigrees ascertained through multiple affected individuals.