Single channel patch clamping and maximum likelihood fitting to kinetic schemes have been used to characterise the adult human muscle nicotinic receptor. Both the wild type receptor and a disease-causing receptor mutation were studied. The mutation underlies a case of the muscle degenerative disease termed slow channel congenital myasthenic syndrome. The mutation at positions L78P is a gain of function mutations that result in extended burst durations compared with wild type receptors. The mutations affects both agonist dissociation and channel gating, despite the fact that the mutation is not within the classical acetylcholine binding site or within the pore region. The choice of kinetic scheme to describe the receptor has also been investigated. A scheme incorporating two different binding sites indicates that the dissociation rates of acetylcholine are different at the two binding sites of the human receptor. The effect of using an EC50 constraint as opposed to fixing a rate constant was examined, as was the effects of including either an agonist-blocked state or a desensitised state. In schemes examining agonist association and dissociation to and from open channels, it is shown that the rate of dissociation of acetylcholine is negligible compared with the other kinetic parameters involved in channel activation. In all of the schemes tested, consistent values were obtained for the rates of diliganded channel opening, diliganded channel closing, and the total dissociation rate of agonist from the diliganded receptor. In addition the effect of conotoxin-GI on the nicotinic receptor was examined at the single channel level. It is shown that this toxin increases the proportion of monoliganded openings and has asymmetric effects on acetylcholine dissociation from the two binding sites.