Aspects of insulin secretory mechanisms and models of diabetogenic B cell damage are discussed. Measurements of fluxes of 3H-labelled triphenylmethylphosphonium ion, 86Rb+, 42K+, 22Na+, and 45Ca2+ in isolated islets indicate that the triggering of insulin release depends on alterations in the interaction of ions with the B cells. One difficulty in the detailed analysis of these alterations are uncertainties which arise when macroscopic concepts for homogenous phases are applied to microscopic and heterogenous compartments, as exemplified by the meaning of pH in insulin secretory granules and of membrane electric potential. Nonetheless, the importance of an apparent decreased K+ permeability in mediating the insulin-releasing action of glucose, and of an apparent increased Na+ permeability in mediating the potentiating action of acetylcholine is emphasized. Fluorescent probing of Ca2+ by chlorotetracycline revealed effects of glucose alone as well as glucose-dependent and atropine-sensitive effects of acetylcholine. Although acetylcholine, sulfonylureas, and certain thiol-blocking agents may stimulate insulin release by direct effects on the B cell plasma membrane, a high capacity for D-glucose transmembrane transport has probably evolved in order that the interior of the B cells can always sense the circulating glucose concentration. A signal to secretion is thought to be transmitted from glucose metabolism to altered ion fluxes by intervention of reduced pyridine nucleotides and hypothetical redox protein for which thioredoxin may be a model. The insulin secretory defect in hereditary diabetic C57BL/KsJ-db/db-mice is apparently linked to a decreased basal permeability for K+ and a failure of the B cells to decrease further this permeability in response to glucose. Functioning B cells are acutely damaged when exposed to heterologous serum or alloxan in vitro; cytotoxic activation of complement by the alternative pathway could perhaps occur during islet inflammation. Protection experiments with free-radical scavengers in vitro and in vivo support the theory that hydroxyl radicals are instrumental in the production of alloxan diabetes. Rapid reduction of alloxan by thioredoxin in the presence of molecular oxygen and NADPH leads to strong chemiluminescence from luminol indicative of an intense radical protection. The sensitivity of B cells to alloxan may be due to physiological specializations of their plasma membranes, involving the highly effective glucose carrier or the hypothetical oxidation/reduction systems or both.