An observational technique, demonstrated to provide reliable data under a variety of conditions, was used to evaluate the effects of d-fenfluramine, l-fenfluramine, d-amphetamine, and fenfluramine-amphetamine combinations on eating behaviour in human subjects. Following an overnight fast, subjects ate lunch from a dispenser allowing free access to a choice of ten sweet and savoury foods, of varying macronutrient composition. D-fenfluramine (30mg) and d-amphetamine (15mg) reduced food intake; l-fenfluramine (30mg) was essentially inactive. The net effects of d-fenfluramine and d-amphetamine on food intake were additive, but the behavioural mechanisms of action were different for the two drugs. D-amphetamine decreased the duration of the meal, as well as the time spent chewing or manipulating food and the number of bites, but had no effect on eating rate; by contrast, d-fenfluramine decreased the rate of eating, but had no effect on meal duration.