Human volunteer subjects of normal weight received oral doses of (+)amphetamine (10 mg) or (±)fenfluramine (30 mg and 60 mg) together with a placebo control according to a within-subjects design. The effects of these treatments were monitored by measuring food intake in a test meal, subjective ratings of hunger motivation and the micro-structure of eating behaviour abstracted from videotaped recordings of the test meal. Various measures of the rate of feeding were computed from these recordings. Amphetamine and fenfluramine (60 mg) showed generally similar effects on food intake and on the subjective experience of hunger, but displayed differing actions on the fine structure of eating. Amphetamine increased latency to initiation of eating and increased the rate of food ingestion, whilst fenfluramine slowed the local rate of eating and eliminated the characteristic decline in the rate of feeding across the course of a meal. These findings display certain resemblance to the results of animal experiments involving similar pharmacological manipulations and emphasise the importance of measuring rate of feeding in animal and human studies. The results of this study suggest that the micro-analysis of feeding behaviour not only provides a tool for understanding systems involved in the modulation of food consumption but also reveals information which may be helpful for the use of drugs in the treatment of obesity.