Young growing rats weighing approximately 83 g were fed on diets containing kidney bean (Phaseolus vulgaris) lectin (phytohaemagglutinin, PHA) in the range of 0-0.45 g/kg body weight for 10 d to ascertain whether there was a minimum dose below which the lectin had no significant effect on body and skeletal muscle weights in comparison with pair-fed lectin-free controls. Averaged over all experiments, PHA doses of less than 10 mg/d (0.12 g/kg body weight) reduced body dry by 1.14 (SE 0.25) g when compared with controls. Between 10 and 27 mg/d (0.12-0.32 g/kg body weight) a further reduction of 0.64 (SE 0.21) g occurred, suggesting a slight but steady decline of body dry weight with increasing dose. However, above 27 mg/d the depression of growth and changes in body composition accelerated. The difference between the proportional losses of skeletal muscle and body weight was not significant at doses of PHA below 10 mg/d (0.12 g/kg body weight) but the ratio of these losses rose to 1.5-2.0 at doses above this. The proportional decrease in lipid weight exceeded that of both body and skeletal muscle weights at all lectin doses, suggesting that lipid catabolism was the first target of the PHA effect. Plasma insulin level was depressed at the PHA dose of 0.02 g/kg body weight at which growth depression and muscle atrophy were minimal but, contrary to expectations, plasma glucose levels remained stable over the whole PHA dose range. It appears that despite a PHA-induced lowering of blood insulin, glucose catabolism is elevated by an unknown, possibly hormonal, compensatory mechanism. Thus, because low insulin levels facilitate the mobilization and catabolism of lipids, it may be possible to use low doses of PHA to reduce hyperglycaemia and body fat.