It has been reported that caffeine (1.5-30 mg/kg i.p.) as well as specific A1 (DPCPX, 8-cyclopentyl-1,3-dipropylxanthine) receptor antagonists fail to increase extracellular dopamine (DA) in the shell of the nucleus accumbens (NAc). However, it has also been reported that caffeine (10 and 30 mg/kg i.p.) and the A1 antagonist 8-cyclopentyl-1,3-dimethylxanthine (CPT) increases NAc shell DA. To clarify this issue rats were implanted with microdialysis probes at different sites in the NAc shell, in the medial prefrontal cortex (PFCX, infralimbic cortex), and at the border between those areas. Irrespective of probe placement within the NAc shell and of the use of different surgical anesthetics (chloral hydrate and ketamine), we failed to observe changes in dialysate DA after 10 and 30 mg/kg i.p. of caffeine. Similarly negative results were obtained with DPCPX and CPFPX, two potent and selective A1 receptor antagonists. A significant increase of DA was obtained after caffeine when probes were located at the border between the NAc shell and the PFCX (10 and 30 mg/kg) or in the PFCX (10 mg/kg). In view of this and of our previous report that caffeine increases dialysate DA in the medial PFCX, we conclude that the increase in dialysate DA by caffeine observed by others arises from the medial PFCX rather than from the NAc shell as a result of placement of microdialysis probes at the border between the NAc shell and the PFCX.