Mechanisms of anaesthetics still remain unclear. However, various attempts have been made to elucidate the effects of anaesthetics at the molecular level. The nicotinic acetylcholine receptor has proved to be a good model of membrane-bound ligand-gated ion channels. It is available in abundance from Torpedo electroplaques, thus enabling multiple experimental approaches. The receptor exists in different states: the resting state, the open state and the desensitized state, amongst others. For each of these states, effects of general anaesthetics at the receptor molecule have been shown. Displacement studies show barbiturates to bind to a site on the resting state of the nicotinic acetylcholine receptor; cation flux studies suggest barbiturates may also bind to the same, or a similar, site on the open state of the receptor when inhibiting its function. Long-chain alcohols inhibit the open receptor, perhaps by binding to such a site or sites. However, short-chain alcohols do not inhibit and do not share this long-chain alcohol binding site; instead they nonspecifically enhance the agonist's apparent affinity. All the alcohols also cause desensitization by a non-specific mechanism possibly involving perturbation of the lipid bilayer. Thus, general anaesthetics exert both specific and non-specific actions on the acetylcholine receptor.