1. Experiments were carried out to quantify the stimulation-evoked overflow of catecholamines and purines (ATP, ADP, AMP and adenosine) from an in vitro sympathetic nerve-smooth muscle preparation of the guinea-pig vas deferens and from isolated bovine adrenal chromaffin cells. The superfused preparations were stimulated for 60 s with electrical field stimulation (EFS; vas deferens), dimethylphenylpiperazinium (chromaffin cells) or KCl (both preparations). 2. Samples of superfusate were taken at 10 s intervals during the 60 s stimulation period for analysis of purines by HPLC-fluorescence detection and catecholamines by HPLC-electrochemical detection. 3. The evoked overflow of catecholamines and purines from chromaffin cells occurred with the same time course and in a constant ratio of approximately 4:1 (catecholamine to purine). These findings are compatible with the release of catecholamines and purines from a homogeneous population of exocytotic vesicles in the chromaffin cells. 4. The evoked overflow of purines and noradrenaline (NA) from the vas deferens preparation differed from the pattern of overflow from chromaffin cells and there was also some temporal disparity in the overflow of the two cotransmitters. The evoked overflow of ATP exceeded that of NA. In addition, the overflow of NA was tonic while the overflow of ATP and the other purines was phasic. 5. The EFS-evoked overflow of NA and the purines from the guniea-pig vas deferens preparation was examined after treatment with the neuronal amine-uptake inhibitors desipramine and cocaine, the alpha 1-adrenoceptor agonist methoxamine, the alpha 1-adrenoceptor antagonist prazosin, the alpha 2-adrenoceptor antagonists idazoxan and yohimbine, the noradrenaline-depleting drug reserpine and the adrenergic neuron-blocking agent guanethidine. The results of these studies, together with an analysis of the metabolic degradation of extracellular ATP, indicated that the temporal disparity in the overflow of NA and ATP is unlikely to be due to differences in the clearance of the cotransmitters or to the release of purines from non-neuronal sites. These results indicate that evoked overflow of the cotransmitters accurately reflects release from nerves. This pattern of release from nerves suggests that the two cotransmitters are released from two separate populations of exocytotic vesicles. 6. Superfusion of the vas deferens with exogenous epsilon-ATP, a fluorescent derivative of ATP, revealed that there was essentially no metabolism of the nucleotide over 60 s unless the tissue was subjected to EFS. Upon EFS, there was a rapid and nearly complete degradation of ATP with a corresponding increase in ADP, AMP and adenosine. This indicates the presence of a nerve stimulation-dependent metabolism of ATP.