Correction to: Nature https://doi.org/10.1038/nature14886 Published online 5 August 2015

In this Article, the configuration of BU72 at the benzylic amine was mistakenly reported as (S). Re-fitting this stereocentre to (R) improved the fit of the observed density in the X-ray structure. The (S)-configuration was originally assigned on the basis of the expected stereochemistry from ref. 1. The origin of the inversion is unclear, but we propose that the original assignment, based only on nuclear Overhauser effect experiments, may well have been incorrect. We thank Thomas Munro for bringing this error to our attention. The Protein Data Bank (PDB) entry of the active μ-opoid receptor bound to the BU72 agonist has been updated as 5C1M. The original Article has not been corrected online.