The proposed project will integrate original survey data with university administrative data to support analyses that address the ways in which STEM doctoral student’s individual characteristics and training environments shape their acquisition of skills and the effects skill acquisition has on scientific and employment outcomes after graduation. We assess skill acquisition using a variety of validated survey instruments to measure analytic skills (data interpretation, analytic thinking) and softer skills such as management and leadership. Administrative data for more than 30 universities maintained by the Institute for Research on Innovation and Science (IRIS) provides detailed information on the research teams and collaboration networks in which STEM doctoral students receive the bulk of their training. Linkages between IRIS data and restricted federal data sets maintained by the U.S. Census Bureau provide comprehensive information about doctoral recipients’ employment outcomes and earnings. Linkages between IRIS data and research productivity data including patents, dissertations (ProQuest), and publications (PubMeb and Web of Science) provide multiple sources of information about their scientific outcomes. The research proposed here will provide new insight into the opportunities and barriers to participation and success in STEM graduate training and into the relationship between key features of graduate training and individual level outcomes. In so doing this work will directly address recommendations made in numerous recent reports by National Academy of Science panels that address graduate STEM training and workforce development.