To evaluate whether cyclic nucleotides play a role as mediators in antral gastrin release, the following in vivo experiments were performed in dogs. An antral pouch was constructed and the rest of the stomach, the pancreas and small and large intestine were resected. A gastric artery supplying the pouch was cannulated, and after a basal period with saline infusion either dibutyryl cyclic 3',5'-adenosine monophosphate (cyclic AMP, 0.2 mg per kg per min), dibutyryl cyclic 3',5'-guanosine monophosphate (cyclic GMP, 0.05 mg per kg per min) or saline were infused over a period of 60 min. To prove the viability of the pouch and to show its ability to release gastrin with proper stimulation, bethanechol chloride (urecholine) was infused into the gastric artery at the end of the experiment. Blood samples were taken from the portal vein and assayed for gastrin. Neither cyclic AMP nor cyclic GMP infusion was found to increase portal gastrin concentration to a significant degree. A marked increase in portal gastrin concentration however, was observed when bethanechol chloride was infused. These studies lend no support to the thesis that cyclic nucleotides mediate gastrin release in the dog.