An analysis of nearly 24,000 grant applications at the Canadian Institutes of Health Research (CIHR) finds that women are less successful in receiving funding if reviewers are explicitly asked to review the principal investigator, rather than when they are asked to assess the quality of the science.