In order to test the hypothesis that size of place of residence and urban proximity preferences constitute factors in population dispersal migration behavior, a random sample of 777 Pennsylvania households plus a sample screened for moving probability (N=319) were surveyed via personal interviews in 1974. A follow-up survey on actual migration patterns was conducted by phone one year later. Data were analyzed in terms of: (1) preferred size of place of residence by size of place of origin (central city or suburb over 500,000; medium-sized city of 50,000-500,000; smaller city/village under 50,000; and countryside); (2) size of place of destination by preferred size of place of residence; (3) percent of movers attaining preferred size of place of destination and zero order and partial correlation coefficients between preferred and size of place of destination; (4) percent of movers attaining preferred destination with respect to a large city and zero order and partial correlation coefficients between preferred and actual destination with respect to a large city. Data did not support the population spread approach (growth spread in rural areas), as most respondents who preferred a smaller size of place also wanted to be within commuting distance of an urban center, nor did the data indicate correlation between residential preferences and actual migration behavior among those preferring smaller rural places. (JC)