BACKGROUND Previous data from this unit suggest that postnatal growth retardation (PGR) is inevitable in preterm infants. However, the study was performed in a single level III neonatal intensive care unit and applicability of the findings to other level III or level I-II special care baby units was uncertain. OBJECTIVE To examine postnatal hospital growth and to compare growth outcome in preterm infants discharged from four level III tertiary care units and 10 level I-II special care baby units in the former Northern Region of the United Kingdom. METHODS Preterm infants (< or = 32 weeks gestation; < or = 1500 g) surviving to discharge were studied. Infants were weighed at birth and discharge. Body weight was converted into a z score using the British Foundation Growth Standards. To ascertain the degree of PGR, the z score at birth was subtracted from the z score at discharge. Data were evaluated using a combination of split plot (level III v I-II=main factor; individual centre=subfactor) and stepwise regression analyses. Results were considered significant at p < 0.05. RESULTS A total of 659 (level III, n = 335; level I-II, n = 324) infants were admitted over a 24 month period (January 1998-December 1999). No differences were detected in birth characteristics, CRIB score (a measure of illness in the first 24 hours of life), length of hospital stay, weight gain, weight at discharge, or degree of PGR between infants discharged from level III and level I-II units. Significant variation was noted in length of hospital stay (approximately 35%; p < 0.001), weight gain (approximately 33%; p < 0.001), weight at discharge (approximately 20%; p < 0.001), and degree of PGR (approximately 200%; p < 0.001) between the level III units. Even greater variability was noted in the duration of hospital stay (approximately 40%; p < 0.001), weight gain (approximately 60%; p < 0.001), weight at discharge (approximately 40%, p < 0.001), and degree of PGR (approximately 300%, p < 0.001) between the level I-II units. CONCLUSIONS These data stress the variable but universal nature of PGR in preterm infants discharged from level III and I-II neonatal intensive care units and raise important questions about nutritional support of these infants before and after hospital discharge.