The role of glucose in the regulation of plasma cholecystokinin (CCK) level was investigated in healthy control subjects and patients with non-insulin-dependent diabetes mellitus (NIDDM). Plasma CCK concentration was determined by a specific and sensitive bioassay and by a highly sensitive and reliable double-antibody radioimmunoassay using OAL-656 as an antiserum. In control subjects, ingestion of Trelan G-75 (1,200 mOsm/L, 225 mL), which is equivalent to 75 g glucose as metabolic products, caused a rapid and significant increase in plasma CCK bioactivity from 1.3 ± 0.2 to a peak of 5.8 ± 0.6 pmol/L and immunoreactive CCK concentration from 1.2 ± 0.1 to 4.6 ± 0.6 pmol/L. Ingestion of 75 g glucose in 225 mL water (33.3% solution) increased plasma CCK bioactivity to a similar degree to that observed following Trelan G-75 (peak responses, 4.5 ± 0.4 pmol/L). The same volume of 0.9% NaCl solution or water failed to increase plasma CCK concentration. A smaller dose of glucose (50 g/150 mL water) increased plasma CCK concentration, although the peak level (3.0 ± 0.5 pmol/L) was less than that observed following 75 g glucose. In patients with NIDDM, Trelan G-75 ingestion increased CCK concentration, but the peak level was lower, albeit insignificantly, than that of normal subjects. When the maximal increment of plasma CCK above the basal value was compared between control and NIDDM subjects, the differences were statistically significant (NIDDM, 3.6 ± 0.1 pmol/L; control, 5.0 ± 0.4; P < .01). However, integrated CCK responses to Trelan G-75 in NIDDM (165.8 ± 15.5 pmol/120 min) were not significantly different from those in control subjects (189.8 ± 15.9 pmol/120 min). Peak CCK bioactivity occurred within 10 to 30 minutes of ingestion, preceding the increase in glucose and insulin. These results suggest a possible effect of CCK on insulin release in humans, and that the CCK secretory response to glucose in well-controlled diabetic patients is not significantly altered.