Nature Neuroscience, Published online: 23 April 2026; doi:10.1038/s41593-026-02288-y

A recent study by Epp et al. uses advanced, quantitative functional MRI measures to demonstrate that the ‘canonical’ interpretation of blood-oxygenation-level-dependent (BOLD) functional MRI (fMRI) — that increases and decreases in brain activation are accompanied by corresponding changes in blood flow and oxygen metabolism — does not strictly hold across the human brain. Although the authors provided a balanced interpretation, this has been viewed by others as undermining fMRI. We discuss whether the findings bring into question the validity of fMRI-based measures of brain function.