Glyceraldehyde (GA) has been used to study insulin secretion for decades and it is widely assumed that beta-cell metabolism of GA after its phosphorylation by triokinase is similar to metabolism of glucose; that is metabolism through distal glycolysis and oxidation in mitochondria. New data supported by existing information indicate that this is true for only a small amount of GA's metabolism and also suggest why GA is toxic. GA is metabolized at 10-20% the rate of glucose in pancreatic islets, even though GA is a more potent insulin secretagogue. GA also inhibits glucose metabolism to CO2 out of proportion to its ability to replace glucose as a fuel. This study is the first to measure methylglyoxal (MG) in beta-cells and shows that GA causes large increases in MG in INS-1 cells and d-lactate in islets but MG does not mediate GA-induced insulin release. GA severely lowers NAD(P) and increases NAD(P)H in islets. High NADH combined with GA's metabolism to CO2 may initially hyperstimulate insulin release, but a low cytosolic NAD/NADH ratio will block glycolysis at glyceraldehyde phosphate (GAP) dehydrogenase and divert GAP toward MG and D-lactate formation. Accumulation of D-lactate and 1-phosphoglycerate may explain why GA makes the beta-cell acidic. Reduction of both GA and MG by abundant beta-cell aldehyde reductases will lower the cytosolic NADPH/NADP ratio, which is normally high.