This paper discusses a study to determine patterns of residential segregation by sex in Washington, D.C. over a 30 year period. Objectives were to explain statistically sex ratio distribution in housing, identify unsuspected sex discrimination processes affecting urban populations, and to clarify residential clustering. Washington, D.C. was selected for study because it has been exceptionally female dominant since 1940. Data were obtained from earlier studies of housing patterns and from census figures. The method was to analyze statistically all census tracts within the central city in 1970 and to compare results with data from earlier studies of housing patterns since 1940. Variables tested in addition to sex-specific factors included racial composition of city tracts, local economic or institutional factors such as military installations, age structure, and median per capita income of tract residents. Findings indicated that females were numerically dominant in a zone of contiguous tracts within the central city; this dominance persisted for all census years studied; male dominant tracts were widely scattered in contrast to the clustered female dominant tracts; female dominant tracts were characterized by greater proportions of their population over 60 years of age and more owner-occupied housing; and segregation by sex is less than, and seemingly unrelated to, segregation by income, race, or ethnic group. The conclusion is that female dominant census tract populations in Washington, D.C. persist because of sex-specific factors including the desire to live in well-settled areas with high percentages of owner-occupied housing and to settle in areas where there are large numbers of other women. (DB)