The kinetics of zinc release and insulin secretion were compared in perifused islets previously loaded for 24 h with 65Zn. In the absence of detectable insulin secretion, fractional basal 65Zn release was constant at 3-5%/h throughout the experimental periods, suggesting that basal efflux of zinc is primarily from a single extragranular compartment. Glucose (25 mM) alone caused a prompt release of both 65Zn and insulin. Previous studies using cell fractionation showed that approximately one third of islet 65Zn is in the granular fraction. Thus, if zinc efflux represents only that being cosecreted from the granule, a 65Zn to insulin fractional secretion ratio of approximately 0.33 would be expected. However, the ratio of 65Zn to insulin fractional secretion was only 0.16, indicating that glucose not only causes cosecretion of zinc associated with stored insulin, but also independently decreases 65Zn efflux from extragranular sources. At low glucose concentrations, 3-isobutyl-1-methylxanthine (IBMX; 1 mM) did not affect either insulin or 65Zn efflux; however, the combination of glucose plus IBMX caused a large and parallel biphasic release of both 65Zn and insulin at ratios approximating 0.33. Similar results were obtained with glucose plus forskolin. Leucine (20 mM) plus IBMX caused a characteristic rapid secretion of insulin which waned with time. Efflux of 65Zn was also prompt; however, the ratio of secreted 65Zn to insulin gradually increased during stimulation from a near-theoretical value for cosecretion at the initial peak to 0.70 by the end of the stimulation. 65Zn efflux was sustained even after termination of the leucine plus IBMX stimulus. alpha-Ketoisocaproic acid caused a similar stimulation and waning of insulin secretion, but with a consistent 65Zn to insulin fractional secretion ratio of about 0.33. With both leucine and alpha-ketoisocaproic acid, poststimulatory 65Zn efflux remained elevated. Perifusion of unlabeled zinc (1 mM; a concentration that inhibits insulin secretion) caused prompt and sustained release of 65Zn; efflux remained high at least 30 min after termination of zinc administration. Results expand previous batch islet experiments to show that zinc and insulin can be temporally cosecreted at ratios quantitatively consistent with their storage ratio in the granules without conservation or intracellular reutilization. In addition, there is 65Zn efflux from nongranular compartments which is stimulated or inhibited, depending on the secretagogue.(ABSTRACT TRUNCATED AT 400 WORDS)