Pre-clinical animal research is typically based on single laboratory studies conducted under highly standardized conditions. But in a new study, researchers show that this near-universal practice may actually help to explain the poor reproducibility of pre-clinical animal research. Instead of standardized conditions, diversity may be better.