The proinsulin-like component comprised approximately 20% of total circulating basal immunoreactive insulin in 15 patients without islet cell tumors. 15 min after oral glucose, the concentration of the proinsulin-like component was unchanged and its percentage of the total immunoreactive insulin decreased with the acute release of the insulin component. By 2 hr after oral glucose, the concentration of the proinsulin-like component increased and the insulin component concentration decreased so that the percentage of the proinsulin-like component was essentially the same as in the basal state. In five patients with islet cell tumors and fasting hypoglycemia, basal proinsulin-like component ranged from 26 to 79% of the total immunoreactive insulin. While basal proinsulin-like component was higher in the islet cell tumor patients, the fluctuations after stimulation were qualitatively similar to the nontumor patients. Acute stimulation with glucose, tolbutamide, leucine, and streptozotocin mainly released the insulin component resulting in a fall in the per cent proinsulin-like component with a subsequent increase in percentage of this component as the total insulin concentration returns towards basal levels. Three islet-cell tumor patients with less than 46% proinsulin-like component had favorable therapeutic responses to diazoxide whereas one patient with over 80% proinsulin-like component was completely refractory. Syrian hamsters bearing islet cell tumors provided an excellent model for islet cell tumors in man. These animals have a high proportion of a proinsulin-like component in plasma; stimulation of tumor slices in vitro with tolbutamide and glucagon releases mainly the insulin component similar to the observations in man. These studies suggest that the mechanisms regulating the release of the proinsulin-like and of the insulin components are different.