Caffeine produces effects on cognitive function particularly relating to aspects of attention such as reaction time. Considering the plasma exposure levels following regular caffeine intake, and the affinity of caffeine for known protein targets, these effects are likely mediated by either the adenosine A(1) or A(2A) receptor. In the present studies, two rat strains [Long-Evans (LE) and CD] were trained to asymptote performance in a test of selective attention, the 5-choice serial reaction time task (5-CSRTT). Next, the effects of caffeine were compared to the selective A(2A) antagonists, SCH 412348 and KW-6002 (Istradefylline), and the A(1) antagonist, DPCPX. Further studies compared the psychostimulant effects of each drug. Finally, we tested the A(2A) agonist, CGS-21680, on 5-CSRTT performance and given the antipsychotic potential of this drug class, studied the interaction between CGS-21680 and amphetamine in this task. Caffeine (3-10mg/kg IP) increased reaction time in both LE and CD rats, with no effect on accuracy, an effect replicated by SCH 412348 (0.1-1mg/kg PO) and KW-6002 (1-3mg/kg PO), but not DPCPX (3-30 mg/kg PO). At least with SCH 412348, these effects were at doses that were not overtly psychostimulant. In contrast, CGS-21680 (0.03-0. 3mg/kg IP) slowed reaction speed and increased omissions. Interestingly, at a comparatively low dose of 0.03 mg/kg, CGS-21680 attenuated the increased premature responding produced by amphetamine (1mg/kg IP). The present results suggest that the attention-enhancing effects of caffeine are mediated through A(2A) receptor blockade, and selective A(2A) receptor antagonists may have potential as therapies for attention-related disorders. Furthermore, the improvement in response control in amphetamine-treated rats following CGS-21680 pretreatment supports the view that A(2A) agonists have potential as novel antipsychotics.