1. Cat spleens were perfused with Krebs-bicarbonate solution, using a constant-flow pump at a rate of about 7 ml/min at 33-35 degrees C. Noradrenaline (NA) overflow by nerve stimulation at 10 Hz for 20 s was determined with or without flow-stop before and after treatment with cocaine, phentolamine or phenoxybenzamine. In order to determine the effect of flow-stop on overflow, the arterial and the venous flows were occluded by clamping the inflow and outflow tubes during the period of stimulation plus 30, 60 or 120 seconds.2. Without flow-stop, NA output was 0.93+/-0.25 ng/stimulus, which was significantly increased after cocaine (123+/-6.6%), phentolamine (415+/-93%) and phenoxybenzamine (578+/-107%). Phentolamine and phenoxybenzamine were much more effective than cocaine in enhancing overflow.3. Before treatment with drugs, flow-stops of 30, 60 and 120 s reduced NA outputs to 70+/-6.6, 27.5+/-2 and 7%, respectively, of the control outputs without flow-stop. None of the drugs significantly influenced the percentage reductions in NA outputs during a 30 s flow-stop. However, the percentage outputs after cocaine or phenoxybenzamine treatment during a 60 s flow-stop significantly increased to 45+/-2.5% and 57+/-6%, respectively, as compared to the percentage output of 27.5+/-2% from untreated spleens during a corresponding flow-stop period. During flow-stop, there was no appreciable metabolism of the released transmitter.4. Diffusion of the released transmitter from the site of liberation plays only a minor role in the removal of the released NA.5. It is suggested that the NA released by nerve stimulation acts on the presynaptic alpha sites to inhibit its own release by a negative feedback mechanism. Adrenoceptor blocking agents enhance the NA overflow from spleen because they remove this autoinhibition by blocking the presynaptic alpha sites.