The McCollough Effect (ME) is an orientation contingent colour after-effect which has been proposed as an indicator of central neurotransmitter activity. Shute (1979) suggested that the ME could reflect a hippocampal "forgetting" mechanism which should be inhibited by GABAergic neurones and stimulated by cholinergic neurones. The purpose of the present study was to demonstrate that the ME is in fact sensitive to cholinergic and anticholinergic drugs and to compare its sensitivity to more conventional tests of psychomotor and cognitive function. Ten healthy subjects received single doses of physostigmine (0.75 mg SC), hyoscine (1.2 mg), temazepam (20 mg), flecainide (200 mg) or placebo in a double-blind double-dummy presentation. Subjects were tested on a battery of psychomotor and cognitive function tests at baseline and 1 h, and adapted to the ME at 1.5 h. Visual analogue rating scales and conventional tests of psychomotor function and saccadic eye movements indicated that both subjective and objective measures of arousal were impaired by temazepam. The subjective, but not the objective, measures of arousal were also impaired by both hyoscine and physostigmine, but not by flecainide. Initial strength and duration of the ME were decreased by physostigmine and increased by hyoscine and temazepam, relative to placebo (P less than 0.01). Thus, the ME is capable of detecting cholinergic, anticholinergic and GABA mimetic drug effects in man, in therapeutic doses.