1. An insulin-producing cell line, RINm5F, derived from a rat insulinoma was studied. 2. The cellular content of immunoreactive insulin was 0.19 pg/cell, which represents approx. 1% of the insulin content of native rat beta-cells, whereas that of immunoreactive glucagon and somatostatin was five to six orders of magnitude less than that of native alpha- or delta-cells respectively. 3. RINm5F cells released 7-12% of their cellular immunoreactive-insulin content at 2.8 mM-glucose during 60 min in Krebs-Ringer bicarbonate buffer. 4. Glucose utilization was increased by raising glucose from 2.8 to 16.7 mM. There was, however, no stimulation of immunoreactive-insulin release even when glucose was increased from 2.8 to 33.4 mM. A small stimulation of release was, however, found when glucose was raised from 0 to 2.8 mM. 5. Glyceraldehyde stimulated the release of immunoreactive insulin in a dose-dependent manner. 6. At 20 mM, leucine or arginine stimulated release at 2.8 mM-glucose. 7. Raising intracellular cyclic AMP by glucagon or 3-isobutyl-1-methylxanthine stimulated release at 2.8 mM-glucose with no additional stimulation at 16.7 mM-glucose. 8. Stimulation of immunoreactive-insulin release by K+ was dose-related between 2 and 30 mM. Another depolarizing agent, ouabain, also stimulated release. 9. Adrenaline (epinephrine) inhibited both basal (2.8 mM-glucose) release and that stimulated by 30 mM-K+. 10. Raising Ca2+ from 1 to 3 mM stimulated immunoreactive-insulin release, whereas a decrease from 1 to 0.3 or to 0.1 mM-Ca2+ lowered the release. 11. These findings could reflect a relatively specific impairment in glucose handling by RINm5F cells, contrasting with the preserved response to other modulators of insulin release.