SUMMARY The ponderal index (PI) (g/cm3 x 100) was used to define the state of nutrition at birth of 47 light-for-date (LFD) term infants and to determine how intrauterine undernutrition influences growth during the first 6 months of postnatal life, at four age intervals: birth-4, 5-8, 9-12, 13-26 weeks. With the exception of one baby, each PI was <50th centile on the Miller and Hassanein standards: in 24 (51 %) babies the PI was <3rd centile. This suggests that, generally, LFD infants are underweight for length, and by inference, are likely to have experienced intrauterine undernutrition. In the first month the 24 'wasted' infants (PI <3rd centile) gained more rapidly in weight, and grew more quickly in head circumference and length than the 16 'nonwasted' infants (PI >10th centile). Thereafter growth rates were similar. With the exception of weight in the 'nonwasted' infants during the first month, rates of growth (weight, length, and head circumference) in both groups of babies in each of the 3 months after birth were greater than in normal infants. The first 3 months after birth can therefore be defined as the period of 'catch-up' growth in LFD term infants.From their outward appearance light-for-date (LFD) infants who show no signs of congenital malformations or intrauterine infection can be differentiated according to the extent to which they show external signs of intrauterine undernutrition. Some show unmistakable features of wasting while others seem to show none, resembling instead 'perfect miniatures'. Little is known about postnatal growth after intrauterine undernutrition as no study of growth in LFD infants (for example Babson, 1970;Fitzhardinge and Steven, 1972;Brandt, 1976;Martell et al., 1978) congenital malformation or any external feature of intrauterine infection. The weight of each infant was <5th centile for gestation, allowing for sex, mother's height, and birth order (Tanner and Thomson, 1970). Weight, length, and head circumference were measured (D.P.D.) between 24 and 48 hours of age, using methods previously described (Davies, 1976). For each infant the P1 (g/cm3 x 100) was calculated and according to the distribution on the centile standards of Miller and Hassanein (1971) (birth-4, 5-8, 9-12, 13-26