Gut hormones which are proposed as natural satiety factors may act through malaise to suppress short-term feeding instead of through a normal physiological process. Evaluation of the ability of candidate satiety hormones to serve as the unconditioned stimulus in the development of a conditioned taste aversion is not adequate evidence of malaise since agents which do not produce malaise in humans are also capable of conditioning aversions. Examination of the patterns of eating behavior when drugs are given during a meal in rats may serve as a better test of the potential of an agent to produce illness. To evaluate this approach, LiCl (a known aversive agent) was infused into five rats at the start of each free-feeding meal at a dose of 1.9 mg/meal/rat for four days and at 3.8 mg/meal/rat for an additional four days. Feeding patterns during drug infusion were compared with patterns for six days when no drugs were infused. LiCl infusion produced a significant dose-dependent reduction in feeding frequency while the size and duration of meals were not changed. At the higher dose of lithium, daily meal number was reduced from a normal 13.2 +/- 1.4 to 9.9 +/- 0.5 meals/day. This pattern of behavior is significantly different from the behavior exhibited when cholecystokinin octapeptide (CCK-8) was infused using the same paradigm. CCK-8 significantly reduced meal size while meal frequency increased to compensate for the smaller meals. These findings suggest that LiCl infusion during a meal will lead to an aversion to feeding while CCK-8 is not effective at producing an aversion.(ABSTRACT TRUNCATED AT 250 WORDS)