Direct comparisons between radiolabelled and endogenous dopamine (DA) release from superfused rat brain slices have been made. Striatal slices were prelabelled with [3H]dopamine ([3H]DA), then superfused at 0.5 ml/min and the released catecholamines analyzed by HPLC with electrochemical detection and the radioactivity present in superfusate fractions also counted. Two successive 50 mM K+ pulses released similar amounts of endogenous DA from striatal slices, but the second pulse released 50% less [3H]DA than the first. A K+ gradient (5-53 mM) released relatively more [3H]DA compared to endogenous DA at lower K+ than at higher K+ concentrations. Blockade of DA synthesis in vitro by 50 microM a-methyl-p-tyrosine greatly reduced K+-induced endogenous DA release without any major effect on [3H]DA release. Amphetamine (10 microM) greatly increased both basal DA release and release induced by a 5 microM veratrine pulse, but its effects were 3-4 times greater on endogenous than on [3H]DA release. Although a-methyl-p-tyrosine reduced both basal and veratrine-stimulated endogenous DA release from non-prelabelled tissue by over 50% in either the presence or absence of amphetamine, it did not decrease endogenous DA release from prelabelled tissue. These studies indicate that labelled and endogenous amine release do not always occur in parallel, and that major causes of discrepancy between them may include the presence of a large newly-synthesized component in endogenous release and the uneven distribution of labelled amine within endogenous releasable pools. The results also suggest that the prelabelling process itself may alter the pools contributing to subsequent endogenous release. In the light of these studies, the assumption that labelled amine release provides an accurate marker for endogenous release should be reconsidered.