Food intake of four adult male baboons (Papio c. anubis) was monitored during daily experimental sessions lasting 22h. Food was available under a two-component operant schedule. Following completion of the first "procurement component" response requirement, access to food, i.e. a meal, became available under the second "consumption component" during which each response produced a 1-g food pellet. After a 10-min interval in which no response occurred, the consumption component was terminated. A long-acting cholecystokinin (CCK) analog U-67827E (U-67: 0.80-3.2 micrograms/kg) was administered, in the thigh muscle, at 1100 hrs immediately prior to the start of the daily session on Tuesdays and Fridays. U-67 significantly reduced intake during the first 8-h of the session, and intake during the entire 22-h session. The decreased intake was due to a significant decrease in the size of the first meal of the session as a consequence of decreased duration of feeding without a change in response rate. U-67 also produced dose-dependent increases in latency to the first meal of up to 2.5 h. These results demonstrate that a long-acting CCK analog decreases food intake over a prolonged period of time in a naturalistic feeding situation. In addition, the effects of U-67 were limited to the consumption component, suggesting that this CCK analog affected food intake by interacting with physiological mechanisms specifically associated with feeding.