In order to examine the mechanism whereby stunted children have poor develo
pmental levels, we compared the behaviour of stunted (N = 78) and nonstunte
d (N = 26) children aged 12 to 24 months, and examined the relationship of
their behaviour to their developmental levels. The effect of nutritional su
pplementation with or without psychosocial stimulation on the stunted child
ren's behaviour was also examined. The children were observed at home durin
g 4 days over a period of 6 months. The stunted children showed significant
ly more apathy, and less enthusiasm and variety in exploring, were less hap
py and more fussy. Caretakers' vocalisations to them were less warm or inst
ructive. Stunted children's activity level, exploratory and happy behaviour
s were predictive of change in developmental levels measured on the Griffit
hs Scales, from enrolment to 12 and 24 months later. Supplementation predic
ted mental age at 12 and 24 months after enrolment, however, it had no sign
ificant effect on behaviour.