This paper reports on a pilot study of migration to Greater Bombay, initiated on the recommendation of the Population Commission of United Nations, and utilizes both published tables from the 1961 Census of India and a set of specially prepared tables from the same census. Migrants were defined by birthplace and cross-classified by age and duration of residence in Bombay.Data (1901-61) on net migration (obtained from successive age-sex distributions) are analyzed in terms of underlying trends to give historical perspective to the analysis of recent data with special emphasis on changes in industrial and occupotiona structure.For the 1951-61 decade, the extensiveness of out-migration of former in-migrants, its age-sex selectivity, and its high incidence among recent migrants are demonstrated. As is true elsewhere, migration to Bombay is shown to be highly selective for ages of maximum economic activity. Migration streams to Bombay were preponderantly male, and, among males, the married segment predominated. The propensity to migrate was unusually high among minority religious groups. As to educational level, migrants were superior to the general population at origin but inferior to nonmigrants residing in Bombay. The work participation rates of migrants were higher for every age group than for resident nonmigrants; the proportion of employees was higher; and there was evidence of migrant concentration in industries and occupations requiring less skill, less education, and less capital than was true of nonmigrants. There were significant tendencies toward "division of labor" among various migration streams on the basis of skills and abilities acquired not only by formal education but also through tradition and precept. From the standpoint of the promotion of social change, the large volume (and selectivity) of reverse or return migration is especially note-worthy.The paper concludes with a methodological evaluation of the reliability and validity of duration-of-residence data and indicates that the relatively simple techniques of enumeration and tabulation utilized in this pilot study may have wide applicability in other developing countries.