The National Institutes of Health's methods for reviewing and financing academic research proposals are often praised as the gold standard. Some American scientists, though, have recently offered less flattering descriptions, like "broken" and "arbitrary." NIH officials have heard both arguments, and plenty in between, in recent months. They have begun their broadest-ever self-examination of methods for choosing grant proposals for biomedical research. That is important for universities because the NIH is the country's single largest source of money for academic research. Just over half of its budget, $29.2-billion this year, goes to institutions of higher education. The NIH's director, Elias A. Zerhouni, says he is prepared to advocate shaking up the status quo to make improvements. This article reviews some of the obstacles facing the NIH in its grant-making, and some of the proposals for improving the process. (Contains 2 tables and 2 figures.)