ln the late 1970's, the Census Bureau began preparing an experimental set of county population estimates by age, sex, and race. Since 1980, we have released such estimates for July 1, 1982, and 1984, and we are developing estimates for other years. This report briefly describes the procedures used in developing the estimates in the 1980's and indicates changes to the methodology previously used. The original methodology used for the estimates for the 1970's is described in Current Population Reports, Series P-23, No. 103. Those aspects of the procedures which have not changed substantially are not discussed in great detail in this report; in addition, examples of calculations shown in the earlier report are not repeated here.

County estimates by age, sex, and race are prepared by the Bureau of the Census under contract to the National Cancer Institute (NCI). The methodology we use is a cohort-component projection technique in which actual data for the components are substituted for projected values whenever possible and totals are controlled to available estimates by age, sex, race, and geographic area. We consider the estimates experimental because they have not been tested fully against a census and because subnational estimates for race groups, for use as control totals, are only now becoming available.