Genetic analysis of polygenic traits in rats and mice has been very useful for finding the approximate chromosomal locations of the genes causing quantitative phenotypic variation, so-called quantitative trait loci (QTL). Further localization of the causative genes and their ultimate identification has, however, proven to be slow and frustrating. A major technique for gene identification in such models utilizes series of congenic strains with progressively smaller chromosomal segments introgressed from one inbred strain into another inbred strain. Under the assumption that a single causative locus underlies a QTL, nested series of congenic strains were earlier suggested as an appropriate configuration for the congenic strains. It is now known that most QTL are compound, that is, the QTL signal is caused by clusters of loci where alleles exert positive, negative, and interactive effects on the trait in a given strain comparison. It is argued that in this situation an initial series of nonoverlapping contiguous congenic strains over a relatively large chromosomal region will lead to a better appreciation of the underlying complexity of the QTL and therefore more rapid gene identification. Examples from the literature where this strategy would be helpful, as well as a case where it would be potentially counterproductive, are given.