Marker-based segregation analysis (MBSA) is a modification of a published method of combined linkage and segregation analysis (Am J Hum Genet 51: 1111-1126, 1992), to determine whether a candidate gene known to be associated with the disease of interest is truly segregating with the disease in families. Here we outline the conceptual basis of MBSA and present a Monte Carlo method for significance testing. The method is applied to PSORS1, a locus within the major histocompatibility complex (MHC) for which linkage and linkage disequilibrium with psoriasis has already been demonstrated. The results are very consistent with our current knowledge of PSORS1, and suggest that MBSA can provide useful information on genotype-phenotype relationships such as penetrance and allelic heterogeneity.