The effect of 6-aminonicotinamide (6-AN), an antimetabolite mainly of NADP synthesis, on insulin release, was studied. Adults rats were pretreated with 6-AN (35 mg./kg.); six hours later, the in vitro insulin response of isolated islets to glucose was significantly reduced from 561 ± 41 to 175 ± 22μU. of immunoreactive ingiilin (IRI) per five islets per ninety minutes incubation time. On the other hand, in vitro addition of NADP or NADPH (1 or 10 mM) to islets from pretreated rats, restored insulin release. In control islets, addition of NADP or NADPH showed little effect. It is unlikely that 6-AN affected insulin synthesis in vivo, as extractable insulin content of islets did not change (1.5 ± 0.1 vs. 1.4 ± 0.1 mU. IRI per single islet). In vivo studies with 6-AN pretreated rats showed marked hyperglycemia (344 ± 25 vs. 139 ± 11 mg./100 ml.) after a six-hour fast and still significant hyperglycemia following a twenty-four-hour fast (282 ± 33 vs. 94 ± 6). When challenged by intraperitoneal glucose, insulin failed to rise significantly in the 6-AN treated animals. It is thus possible to produce a new variant of experimental diabetes by interfering with NADPH metabolism within the beta cell. Our data suggest that glucose is involved in insulin release as a generator of NADPH via the pentose-phosphate shunt.