Experimental diabetes was produced in cats by partial pancreatectomy using a short and technically simple surgical procedure. Electrocautery was used to cauterize pancreatic blood vessels and seal free edges of remaining pancreatic tissue to prevent secretion of pancreatic enzymes into the peritoneal cavity. In a second group of animals, partial pancreatectomy was followed by local injection of alloxan into an arterial branch of the cranio-mesenteric artery. The combined procedure resulted in diabetes mellitus in 100% (8 of 8) animals as compared to only 70% (14 of 20) in those subjected to partial pancreatectomy alone. In addition, the alloxan-pancreatectomized cats had a reduced latency period prior to onset of chronic hyperglycemia (4.8 days compared to 19.3 days postoperatively in pancreatectomized cats). The diabetic cats were maintained in poor metabolic control (blood glucose approximately 300 mg/dl) by daily injections of low doses of long-acting insulin. Pancreatic enzyme supplementation was given by mouth. Weight changes and blood glucose levels were monitored carefully to maintain the health of the animals while keeping them in poor metabolic control.