In rat pancreatic islets, cysteine analogues, including glutathione, acetylcysteine, cysteamine, D-penicillamine, L-cysteine ethyl ester, and cysteine-potentiated glucose (11.1 mM) induced insulin secretion in a concentration-dependent manner. Their maximal effects were similar and occurred at approximately 0.05, 0.05, 0.1, 0.5, 1.0, 1.0 mM, respectively. At substimulatory glucose levels (2.8 mM), insulin release was not affected by these compounds. In contrast, thiol compounds, structurally different from cysteine and its analogues, such as mesna, tiopronin, meso-2,3-dimercaptosuccinic acid (DMSA), dimercaprol (BAL), beta-thio-D-glucose, as well as those cysteine analogues that lack a free-thiol group, including L-cystine, cystamine, D-penicillamine disulfide, S-carbocysteine, and S-carbamoyl-L-cysteine, did not enhance insulin release at stimulatory glucose levels (11.1 mM); cystine (5 mM) was inhibitory. These in vitro data indicate that among the thiols tested here, only cysteine and its analogues potentiate glucose-induced insulin secretion, whereas thiols that are structurally not related to cysteine do not. This suggests that a cysteine moiety in the molecule is necessary for the insulinotropic effect. For their synergistic action to glucose, the availability of a sulfhydryl group is also a prerequisite. The maximal synergistic action is similar for all cysteine analogues tested, whereas the potency of action is different, suggesting similarity in the mechanism of action but differences in the affinity to the secretory system.