Female Wistar rats were allowed to self-administer nicotine solutions through indwelling jugular vein cannulae for 23 h per day for periods from three to five weeks. Two response levers were available to the rats; responding on one lever, designated the active lever, produced an immediate infusion of nicotine solution or saline. A second lever for which responding had no programmed consequences was introduced as a control for the locomotor stimulant action of low doses of nicotine. Baseline lever response rates were determined over a period of one week, in which active lever responding produced an infusion of saline. Rats were then allowed access to varying doses of nicotine or saline for a further two or three weeks. Response rates on the active lever increased significantly in rats with access to nicotine at a dose of 30 micrograms kg-1 per response. However, control lever response rates were also significantly elevated. The role of nicotine-induced locomotor stimulation in the self-administration behaviour was further evaluated in a dose-reduction experiment, in which the dose of nicotine available to rats responding for 30 micrograms kg-1 per response was reduced to 3 micrograms kg-1 per response. This resulted in a significant differential increase in active lever responding relative to control lever responding. The results suggest that nicotine is positively reinforcing in rats which had not previously been deprived of food or water or received prior drug treatment, but also indicate that nicotine induced locomotor stimulation may contribute to the observed increases in lever response rates when rats self-administer nicotine.