Researchers are impeding reproducibility by not identifying experimental resources or properly sharing data, says a study (N. A. Vasilevsky et al. PeerJ 1, e148; 2013). In the methods sections of 238 biomedical journal articles from 2012–13, 54% of resources such as antibody types and cell lines were not fully defined. A lack of identifiers also plagues online data-sharing, says lead author Nicole Vasilevsky, an ontologist at Oregon Health and Science University in Portland, adding that editors, funding agencies and grant reviewers should mandate resource identification. “If researchers don't keep track of what reagents they used, it could impact their own — and others' — research,” she says.