Altered fitness of affected individuals can substantially influence the pattern of risk of illness in relatives. This altered risk can result from both changes in the distribution of genotypes and changes in the genotype-specific proportion affected or GSPA. This report examines the impact of altered fitness on the GSPA in relatives of affected probands, under a simple genetic model. The model assumes that illness results from a generalized single major locus and that fitness is dependent only upon phenotype. It is shown that i) in progenitors of probands (e.g., parents and grandparents), GSPA can vary widely as a function of the fitness of affected individuals; ii) the change in GSPA is equal in parents and grandparents; and iii) in non-progenitors (e.g., offspring, siblings, uncles/aunts), GSPA is independent of fitness effects. The GSPA in parents and grandparents is a function both of the relative fitness of affected individuals (F) and the penetrance (Q) that can be approximated by the simple formula: GSPA = FQ/(FQ + 1 - Q). The successful application of model-fitting methods for the detection of single gene variation to complex phenotypes that significantly alter fitness may, in certain circumstances, require the incorporation of "fitness effects" into the analytic model.