Background. There is general agreement that the number of U.S. science PhDs being trained far exceeds the number of future academic positions. One suggested approach to this problem is to significantly reduce the number of PhD positions. A counter argument is that students are aware of the limited academic positions but have chosen a PhD track because it opens other, non-academic, opportunities. The latter view requires that students have objective information about what careers options will be available for them. Methods. The scientific careers of the 1992-94 cohort of NIH National Institute of General Medical Sciences (NIGMS) Kirchstein-NRSA F32 postdoctoral fellows (PD) was determined by following their publications (PubMed), grants (NIH and NSF), and faculty and industry positions through 2009. These basic life science PDs receive support through individual grant applications and represent the most successful class of NIH PDs as judged by academic careers and grants. The sex dependence of the career and grant success and the influence of the PD mentor's citation record were also determined. Results. Of the 439 1992-94 NIGMS F32 fellows, the careers of 417 could be determined. Although females had significantly higher rates of dropping out of science (22% females, 9% males) there was no significant difference in the fraction of females that ended up as associate or full professors at research universities (22.8% females, 29.1% for males). More males then females ended up in industry (34% males, 22% females). Although there was no significant correlation between male grant success and their mentor's publication record (h index, citations, publications), there was a significant correlation for females. Females whose mentor's h index was in the top quartile were nearly 3 times as likely to receive a major grant as those whose mentors were in the bottom quartile (38.7% versus 13.3%). Conclusions. Sixteen years after starting their PD, only 9% of males had dropped out of science. More females (28%) have dropped out of science, primarily because fewer went into industry positions. The mentor's publication record does not affect the future grant success of males but it has a dramatic effect on female grant success.