Chronic nicotine infusion has been found in a series of studies in our laboratory to significantly improve choice accuracy of rats in the eight-arm radial maze. The current study was designed to compare the effects of chronic nicotine infusion on working and reference memory in a 16-arm radial maze. Nicotine was administered to female Sprague-Dawley rats at approximately 5 mg/kg per day SC via osmotic minipumps. Controls received saline infusions. Chronic nicotine infusion significantly lowered the number of working memory errors compared to controls, whereas the number of reference memory errors was not significantly affected. The modest nicotine-induced reduction in working memory errors was seen as a main effect over the 4 weeks of infusion, but the dearest effect was seen in weeks 3-4 of nicotine administration. For the 2 weeks after withdrawal, the nicotine effect was no longer evident. Acute D-1 challenges were given with the D-1 agonist dihydrexidine (0, 0.25, 0.5 and 1 mg/kg) and the D-1 antagonist SCH 23390 (0, 0.005, 0.015 and 0.05 mu g/kg) during weeks 3-4 of chronic nicotine administration and weeks 1-2 after withdrawal from nicotine. Dihydrexidine caused a modest dose-related increase in reference memory errors but not working memory errors in the nicotine-treated, but not the control rats. The D-1 antagonist SCH 23390 caused a modest though significant decrease in reference memory errors but not working memory errors in the control, but not the nicotine-treated rats. The behavioral specificity of chronic nicotine infusion was demonstrated with selective improvement in working memory function. Pharmacological interactions were seen with chronic nicotine treatment increasing responsivity to D-1 agonist and decreasing responsivity to a D-1 antagonist with regard to reference memory. The mechanisms of this interaction are still undiscovered.