Purpose of review Amnesia, hypnosis and immobility are essential components of general anaesthesia. This review highlights recent advances in our understanding of how these components are achieved at a molecular level. Recent findings Commonly used volatile anaesthetic agents such as isoflurane or sevoflurane cause immobility by modulating multiple molecular targets predominantly in the spinal cord, including gamma-aminobutyric acid(A) receptors, glycine receptors, glutamate receptors and TREK-1 potassium channels. In contrast, intravenously applied drugs such as propofol or etomidate depress spinal motor reflexes almost exclusively via enhancing gamma-aminobutyric acid(A) receptor function. Studies on knock-in animals showed that etomidate and propofol act via gamma-aminobutyric acid(A) receptors containing beta(3) subunits, whereas gamma-aminobutyric acid(A) receptors including alpha(2) and gamma subunits mediate the myorelaxant properties of diazepam. These findings suggest that a large fraction of gamma-aminobutyric acid(A) receptors in the spinal cord assemble from alpha(2), beta(3) and most probably gamma(2) subunits. The hypnotic actions of etomidate are mediated by beta(3)-containing gamma-aminobutyric acid(A) receptors expressed in the brain. In contrast, gamma-aminobutyric acid(A) receptors harbouring beta(2) subunits produce sedation, but not hypnosis. Furthermore, there is growing evidence that extrasynaptic g-aminobutyric acid(A) receptors in the hippocampus containing alpha(5) subunits contribute to amnesia. Summary Clinical anaesthesia is based on drug actions at multiple anatomical sites in the brain. The finding that amnesia, hypnosis and immobility involve distinct molecular targets opens new avenues for developing improved therapeutic strategies in anaesthesia.