Ways that data from four major longitudinal surveys can be used in postsecondary education policy analysis are discussed: the National Longitudinal Survey of Labor Market Experience (NLS-LM), Project TALENT, the Panel Study of Income Dynamics (PSID), and the National Longitudinal Study of the High School Class of 1972 (NLS-HS). The potential use of NLS-LM, TALENT, and PSID as alternatives for or complements to NLS-HS is also addressed. After describing the data collected by each survey, attention is directed to the possible relevance of the survey to research tracing the progress of youths after high school. The surveys are compared concerning information content, sample sizes, sampling, and the problem of data obsolescence. It is suggested that researchers have ignored much of the policy-relevant information available in these four databases by relying chiefly on static analysis and a fairly narrow range of topics: who goes to college or what kind of college, occupational/earnings outcomes, and factors affecting educational attainment. Ways of expanding the scope of research and analysis are suggested. Technical features and peculiarities of content that are likely to affect the use of the surveys for policy research are also identified. (SW)