1. 1. Insulin release has been studied in cultured fetal pancreatic explants from the rat. This preparation has the following advantages: a high degree of reproducibility over prolonged periods of observation and the ability to release a major part (up to 30%) of total insulin content. 2. 2. Glucose alone induced only a small but significant insuling release which was markedly enhanced by the addition of 10 mM caffeine. The smallest effective concentration of glucose in the presence of caffeine was 0.34 mM. 3. 3. Some other sugars and pentitols were tested at the concentration of ii mM in the presence of caffeine. d-Mannose was about half as active as glucose, followed by d-fructose and d-galactose. d-Ribose, xylitol and ribitol were ineffective. 4. 4. 11 mM pyruvate alone induced some insulin release. Its effectivenes was greatly increased in the presence of 10 mM caffeine or of glucagon (3.2 μ/ml). 5. 5. 3-O-Methyl-d-glucose did not inhibit glucose-induced insulin release in the presence of caffeine, while phlorizin was inhibitory only at relatively high concentrations. d-Mannoheptulose, d-2-deoxyglucose, and to a lesser extent d-glucosamine and N-acety-glucosamine, inhibited glucose-induced release. d-Glucosamine and/or N-acetyl-glucosamine alone or in presence of caffeine were stimulatory. 6. 6. Oligomycin (1 μg/ml) strongly diminished insulin release induced by either glucose or pyruvate in the presence of caffeine. 7. 7. It is suggested that insulin release from our preparation is an energy-dependent process, which requires the intracellular accumulation of both 3′,5′-cyclic AMP and of product, or products, of the intermediary metabolism of glucose and/or pyruvate. © 1969.