The effect of purified cholera toxin on secretory processes of exocrine pancreas has been studied in the isolated, saline-perfused cat pancreas and in incubated pieces of rat pancreas. 2. The toxin evoked a biphasic secretory response from the perfused cat pancreas. An initial small phase, which began within minutes of toxin application, was an artefact due to the presence of NaN3 in the cholera toxin preparation as supplied; it could be entirely reproduced by NaN3 at the concentration expected during toxin stimulation. A second, sustained phase of secretion, due to the action of the toxin proper, began within 30-60 min, increasing in magnitude for many hours and persisting in the absence of toxin. It was accompanied by a parellel rise in tissue cyclic AMP concentration, and could be potentiated by theophylline. 3. The composition of the secretion stimulated by cholera toxin resembled that evoked by secretin; e.g. it contained a high concentration of bicarbonate and only basal amounts of digestive enzymes. 4. Similarly, cholera toxin did not stimulate enzyme secretion by incubated rat pancreas, despite large rises in tissue cyclic AMP concentration. 5. Because cholera toxin has thus far been shown to have no other effect than that of stimulating adenylate cyclase, these observations support the conclusion that cyclic AMP does mediate the electrolyte secretory response of the pancreas to secretin, but offers no evidence that cyclic AMP plays a similar role in the regulation of pancreatic enzyme secretion stimulated by cholecystokinin-pancreozymin or acetylcholine.