Growth data of 61 premature infants who were observed for 29 months were analyzed to determine if there is validity to the concept of catch-up growth. Weight, length, and head circumference were recorded at each visit. For each measurement, a growth index was calculated to facilitate comparison of infants of different ages and sexes. The age of a child at each visit was also corrected for prematurity by subtracting the number of weeks premature from the postnatal chronological age to generate a corrected age. The corrected age was then used to create a corrected growth index for all variables at each visit. Growth data evaluated in the traditional manner (without correction) suggested that catch-up growth occurred. However, growth data for corrected age more closely approximated the growth of full-term children. Graphic impressions were generally substantiated by Hotelling's T2 tests and factorial and one-way repeated-measures analyses of variance with Bonferroni adjustments. The corrected growth data of appropriate-for-gestational-age children approximated normal expected growth, whereas small-for-gestational-age children still demonstrated subnormal growth at 29 months of age, despite correction. In both appropriate-for-gestational-age and small-for-gestational-age children, catch-up growth appears to be a statistical illusion created by charting growth of premature infants using chronological age rather than the more appropriate corrected age. This finding has important implications for the follow-up of premature infants and may affect the diagnostic interpretation of failure to thrive.