This paper evaluates the impact of federally subsidized family planning programs in the United States, from 1969 to 1974, on the reproductive behavior of ever-married women, 15-44 years old. The study began with an experimental design and the random assigment of areas to treatment and control but was completed as a multivariate analysis, treating program input as an interval-scaled independent variable. Using "before" and "after" area sample surveys and patient service data from operating programs, the demographic impact of the program was estimated. The impact detected was nonsignificant; according to our analysis, overall use of physician-administered contraception was not affected by the subsidized programs. The only program-induced change of activity related to reproduction that we recorded was increased use of subsidized services in study areas with high program input and increased use of private physician services where program input was low. These findings need to be interpreted with the caution usually exercised in evaluating imperfect field experiments.