1 Drug-naive rats were tested for horizontal and vertical activity in photocell cages, for up to 80 min starting immediately after a subcutaneous injection of (-)-nicotine bitartrate or 0.9% w/v NaCl solution (saline). Nicotine (0.1 to 0.4 mg kg' base) depressed vertical activity and induced ataxia in the first 20 min, but increased both horizontal and vertical activity later in the session; these actions were dose-dependent. A single intraventricular (i.v.t.) injection of chlorisondamine Cl (2 fig base), a quaternary ganglion-blocking drug, given one to two weeks before testing, blocked the ataxic and stimulant actions of nicotine. 2 The antagonistic actions of chlorisondamine (0.2, 1.0, 5.0 jig i.v.t., single administration) were shown to be dose-dependent. The stimulant actions of nicotine were blocked in a dose-dependent way for the duration of the experiment (5 weeks); nicotine's depressant actions were completely blocked at two weeks but not at five weeks.3 A ganglion-blocking dose of chlorisondamine (0.1 mg kg-'), given subcutaneously (s.c.), failed to reduce the behavioural actions of nicotine, whereas a much higher systemic dose (10 mg kg-l s.c.) was effective for at least five weeks. 4 Chlorisondamine failed to alter the behavioural effects of (+)-amphetamine or apomorphine, while blocking those of nicotine. 5 It is concluded that chlorisondamine antagonizes some of nicotine's central actions in a potent, long-lasting and pharmacologically selective way.