Summary. The isolated perfused rat pancreas was stimulated sequentially with arginine or glucose to analyze the time-dependent modulation of insulin release. A 10-min perfusion with arginine (5.0 mmol/1) induced 75% inhibition of the insulin response to repeated arginine stimulation 10min later. When glucose (8.3 retool/l) was given as two pulses, inhibition of the second insulin response was less pronounced. The inhibitory effect generated by arginine also suppressed the insulin response to glucose (27.7 mmol/l), and this inhibitory effect persisted for over 80 min. Stimulation for 30 min with glucose (27.7 mmol/1) strongly potentiated the insulin responses to a pair of arginine stimuli given 20 min later. However, despite augmented secretion rates, the insulin response to the second arginine pulse was still inhibited by 75%. When insulin secretion was strongly amplified by two 10 min pulses of the synergistic mixture of arginine (5.0mmol/1) and glucose (8.3 mmol/1), there was no inhibition of the second insulin response. If glucose (8.3 mmol/1) was present during the first arginine stimulation only, the response to the second arginine pulse was inhibited as in control experiments. However, when glucose was added to the second arginine pulse only, the inhibition generated by the first arginine pulse did not express itself, insulin release remaining similar to control. We conclude that: (1) short stimulations of the pancreas by arginine or glucose generate long-lasting inhibition of the insulin response to subsequent stimulations; (2) synergistic amplification of the insulin response by addition of glucose to arginine obliterates the inhibition; (3) glucose does not suppress the induction of inhibition, it blocks the expression of the inhibitory signal on insulin secretion; (4) these in vitro findings are in keeping with observations in normal and hyperglycaemic man.Key words: Perfused pancreas, insulin secretion, arginine, synergism, time-dependent inhibition, time-dependent potentiation.Studies in vivo and in vitro have demonstrated that the insulin secretion rate is modified by prior sensitization of the pancreas to various insulin releasers [1][2][3][4][5][6][7][8][9][10]. While some stimulators like glucose may generate either a state of potentiation or inhibition in the islet according to the conditions of exposure [2,3,8], arginine induces only an inhibitory state [3,11].The inhibitory state generated by prior exposure to glucose or arginine, termed time-dependent inhibition (TDI), has so far been described only in vivo. In an accompanying paper [11] we demonstrated that it was absent in moderately or slightly hyperglycaemic man (non-insulin-dependent diabetes and obesity), and could be abolished in normal subjects by the acute induction of hyperglycaemia. Since arginine and glucose administration in vivo lead to complex modifications in the milieu of the B cell, and since plasma levels of insulin give only an indirect measure of the insulin secretion rate, the perfused rat pancreas was used to quantify t...