From a modest beginning in 1935 to an income replacement scheme for workers in commerce and industry, the social security system has grown to cover 90 percent of working Americans. It now pays $78 billion a year in benefits and is obligated to pay $4 trillion in future retirement benefits to workers now covered. With a projected tax rate of 30 percent of gross wages by the year 2050 to meet its obligations under current law, the system, in the author's words, is at ""the most crucial juncture of its forty-year life.""This comprehensive review and analysis, the thirteenth in the Brookings series of Studies in Social Economics, discusses social security in relation to other sources of retirement income and clarifies its financing problem, benefit structure, and ambivalent goals. It deals with two main financing questions. First, will the payroll tax yield the revenue needed to pay benefits as the retired population rises as a fraction of the working population? Second, is the regressive social security tax really the best source of revenue for a system with welfare components such as the minimum benefit and dependents' benefits? Though such benefits are viewed as progressive, they are not paid according to need. They meet neither the income redistribution requirements nor the individual equity goals of the social security program. The author provides a comprehensive account of the benefit structure and clarifies its relation not only to the basic goals of social security but also to the country's three-tiered system of income security. She offers recommendations for eliminating ""an irrational and correctable feature of the benefit formula,"" for lowering the dependency ratio, and, most important, for a definition of goals.