Interactions between angiotensin II and adrenoceptor-mediated effects on peripheral sympathetic neurotransmission were investigated in constant flow blood-perfused canine gracilis muscle in situ, without and with pretreatment by non-competitive alpha-adrenoceptor blockade. Angiotensin converting enzyme (ACE)-inhibition by benazeprilat increased nerve stimulation (2 Hz, 4 min)-evoked noradrenaline (NA) overflow (+ 21 +/- 5%) with alpha-adrenoceptors intact, but reduced NA overflow (- 18 +/- 6%) when alpha-adrenoceptors were blocked. Vasoconstrictor responses were slightly reduced by benazeprilat. Subsequent infusion of angiotensin II (Ang II, 20 and 500 ng kg-1 min-1 i.v., raising arterial concentrations from 0.6 +/- 0.2 pM to 1390 +/- 240 and 25,110 +/- 3980 pM, respectively) failed to increase NA overflow or to enhance stimulation-evoked vasoconstriction. Adrenaline (0.4 nmol kg-1 min-1 i.v.) did not change evoked NA overflow before or after benazeprilat, either with or without alpha-adrenoceptor blockade, despite high concentrations (approximately 10 nM) in arterial plasma. Following benazeprilat, propranolol reduced NA overflow (- 24 +/- 3%) only if the alpha-adrenoceptors were blocked. In conclusion, benazeprilat reduced evoked NA overflow in the presence of alpha-adrenoceptor blockade to a similar degree as previously shown in the presence of neuronal uptake inhibition in this model. However, contrasting to our previous findings, benazeprilat enhanced NA overflow and reduced the post-junctional response to nerve stimulation in the absence of alpha-adrenoceptor blockade. This could be related to bradykinin accumulation during ACE-inhibition, in addition to the reduction of Ang II generation. Our data are not compatible with facilitation of NA release by circulating Ang II even at pharmacological dose levels. Although activation of prejunctional beta-adrenoceptors may facilitate evoked NA overflow in this model, circulating adrenaline is ineffective under physiological conditions even after alpha-adrenoceptor blockade. Also, beta-adrenoceptor-mediated prejunctional effects do not seem to involve Ang II in canine skeletal muscle in vivo.