The development of therapeutic agonists for G protein-coupled receptors (GPCRs) is hampered by the propensity of GPCRs to couple to multiple signalling pathways. This promiscuous coupling leads to numerous downstream cellular effects, some of which are therapeutically undesirable. This is especially the case for adenosine A1 receptors (A1Rs) whose clinical potential is undermined by the sedation and cardiorespiratory depression caused by conventional agonists. We have discovered that the A1R-selective agonist, BnOCPA, is a potent and powerful analgesic but does not cause sedation, bradycardia, hypotension or respiratory depression. This unprecedented discrimination between native A1Rs arises from BnOCPA’s unique and exquisitely biased activation of Gob among the six Gαi/o subtypes, and in the absence of β-arrestin recruitment. BnOCPA thus demonstrates a hitherto unknown Gα-selective activation of the native A1R, sheds new light on the fundamentals of GPCR signalling, and reveals new possibilities for the development of novel therapeutics based on the far-reaching concept of biased agonism.