Nicotine can improve attentional functioning in humans, and a number of studies have recently demonstrated that under specific task conditions, nicotine can also improve attention in the rat. Neuronal nicotinic receptors comprise combinations of alpha(2-9) and beta(2-4) subunits, arranged to form a pentameric receptor, with the principal CNS subtypes currently believed to be alpha(4)beta(2) and a homomeric alpha(7) receptor. In the present studies, we attempted to delineate the particular nicotinic receptor subtype(s) contributing to the effects of nicotine on attention by assessing various nicotinic ligands on performance in the five-choice serial reaction time task (5-CSRTT). In rats performing below criterion (<80% correct, >20% omissions to a 1-s visual stimulus), subchronic dosing with nicotine (0.2 mg/kg sc) and the alpha(4)beta(2) agonist SIB 1765F (5 mg/kg sc) increased correct responding and decreased response latencies across the treatment week; whereas the alpha(7) agonist AR-R 17779 (20 mg/kg sc) was without effect. In subjects meeting the criterion, the competitive high affinity (including alpha(4)beta(2)) nicotine receptor antagonist DHbetaE (1-10 mg/kg sc) and the alpha(7) antagonist methyllycaconitine (MLA: 5-10 mg/kg i.p.) did not disrupt performance, whereas at the highest dose, the non-competitive antagonist mecamylamine (0.3-3 mg/kg sc) decreased accuracy and increased response latencies. These changes bore some similarities to those of pre-feeding and the non-competitive NMDA antagonist dizocilpine (0.03-0.06 mg/kg sc), suggesting that mecamylamine-induced performance disruption may relate to non-nicotinic receptor effects. In subjects chronically treated with nicotine, acute nicotine challenge (0.4 mg/kg sc) significantly increased accuracy whilst having no effect on any other performance measures. Finally, in these same nicotine pre-treated rats, the decrease in latency and increase in premature responses induced by nicotine (0.2 mg/kg sc) to a target stimulus of 150 ms was fully antagonised by DHbetaE (3 mg/kg sc) but not MLA (5 mg/kg i.p.). These results suggest that alpha(7) receptors do not play a role in any of the behavioural effects of nicotine observed in the 5-CSRTT, whereas a high affinity site, perhaps alpha(4)beta(2), is more likely involved.