In the News Focus story “Reshuffling graduate training” (31 July, p. [528][1]), J. Mervis describes Roald Hoffmann's proposal to dramatically shift U.S. federal research funding from principal investigator (PI)–controlled grants to graduate student fellowships. However, Mervis does not mention an important argument for Hoffman's proposal: the importance of funding continuity during the entirety of a student's Ph.D. education. As a former graduate program director, I have observed first-hand that stable funding for the duration of a student's Ph.D. studies can affect both the timely completion of the Ph.D. and the student's likelihood of remaining in graduate school. When students (and their PIs) experience funding gaps, responses vary dramatically, even within a given university campus. Some programs leave the student entirely without funding. Other programs provide bridge funding or a teaching assistantship, possibly of finite duration or with strings attached. Even if the student obtains support from another sponsored research grant, such a switch can have serious implications for the content and direction of the student's studies. As a PI, I would gladly give up some amount of funding in return for a secure 5 years of funding for a promising new Ph.D. student to pursue his or her Ph.D. studies. I was pleased to read that senior NSF officials at least considered such proposals in the past and would encourage both NIH and NSF to explore such mechanisms in the future. [1]: /lookup/doi/10.1126/science.325_528