American prosperity and military superiority cannot be maintained with the current shortage of scientists with advanced degrees. How we arrived at this crisis-the embedding of scientific research at male-dominated universities-is less important than what we do to redress it. Approximately ten percent of full professors in the S.T.E.M. disciplines in the United States, and four percent of full professors in physics and engineering, are women, one of the lowest rates among highly developed nations. Top scientists with African-American, Latino, or American Indian ancestry are barely represented. Ultimately, the solution to this gender imbalance is to recruit more native-born women and underrepresented minorities for senior positions in American science. First, we need to attract more women and minorities to pursue advanced degrees. Equally important are new tools to evaluate scientists throughout their careers to replace the unreliable simple count of publications. It merely measures the number of collaborators of a scientist, where men have an overwhelming advantage. Drawing primarily on the literature in program evaluation, the author presents two proposed metrics that would more accurately represent the research contributions of women scholars.