Glucagonlike peptide-1 (7-36) amide (GLP-1) and its receptors are present in several brain regions and may play a role in the physiological control of feeding. To investigate the effect of GLP-1 on eating in the absence of postingestive food stimuli, rats were implanted with gastric cannulas for sham feeding and lateral ventricular cannulas for infusion of GLP-1. Rats (n = 10) sham fed 0.8 mol/L sucrose for 45 min, beginning 5 min after intracerebroventricular (icv) infusion of 2.5 microL of artificial cerebrospinal fluid with 0-30 microg of GLP-1 . Behaviors were observed each minute using a time-sampling technique. Additionally, lick-by-lick records of the microstructural pattern of sucrose intake were made during the first 15 min of each test for five rats receiving 3 and 10 microg of GLP-1. GLP-1 decreased sham-fed intake by as much as 50%, but GLP-1 did not terminate sham feeding. The frequency of observations of feeding was decreased, but the frequency of resting, the terminal item in the behavioral sequence of postprandial satiety in real feeding rats, did not reliably increase. No abnormal behaviors were observed. Although GLP-I did not affect the latency to begin sham feeding, it significantly reduced the initial rate of licking. GLP-I did not affect the motor aspects of licking, because the interlick intervals within individual bursts of licking or overall lick efficiency were normal. These data suggest that intracerebroventricular infusions of GLP-1 inhibit sham feeding by decreasing the orosensory positive feedback that drives licking, rather than by activating physiological satiating mechanisms or nonspecific mechanisms such as aversion or motor incapacity.