The purpose of this article is to explore the effects of organizational and work-group characteristics on the socialization of new scientists. It focuses on the experiences of graduate students and postdoctoral fellows in science. The authors chose to look at outcomes that reflect behaviors (early productivity) and attitudes (willingness to share research findings) since both likely have an impact on the future attitudes and behavior of individuals once they enter the scientific work force. The first point suggested by the data is that the "local setting matters" in graduate education. For both of the outcome variables, a limited number of indicators of organizational structure and climate predict a relatively robust percentage of the variance. Although the rewards of science, from grants to the Nobel Prize, go to individuals, there is evidence that graduate students and postdoctoral fellows who find themselves in the right kind of work setting may have a leg up in their trajectories toward becoming successful scientists. A second overall finding is that "work group size is positively associated with early productivity." The authors conclude that, in a typical university setting, both graduate and postdoctoral students are better off being in larger laboratories. With respect to early productivity, the authors found that life science graduate students and postdoctoral fellows publish and present more than their chemical engineering peers. In spite of the increasingly cross-disciplinary nature of scientific research, this finding suggests the need to continue to explore underlying disciplinary differences that may make generalizations about graduate education inappropriate. (Contains 4 tables.)