Short stature is not the only problem faced by small for gestational age (SGA) children. Being born SGA has also been associated with lowered intelligence, poor academic performance, low social competence, and behavioral problems. Although GH treatment in short children born SGA can result in a normalization of height during childhood, the effect of GH treatment on intelligence and psychosocial functioning remains to be investigated. We show the longitudinal results of a randomized, double-blind, GH-dose response study initiated in 1991 to follow growth, intelligence quotient (IQ), and psychosocial functioning in SGA children during long-term GH treatment. Patients were assigned to one of two treatment groups (1 or 2 mg GH/m(2) body surface.d, or approximately 0.035 or 0.07 mg/kg.d). Intelligence and psychosocial functioning were evaluated at start of GH treatment (n = 74), after 2 yr of GH treatment (n = 76), and in 2001 (n = 53). IQ was assessed by a short-form Wechsler Intelligence Scale for Children-Revised or Wechsler Adult Intelligence Scale (Block-design and Vocabulary subtests). Behavioral problems were measured by the Achenbach Child Behavior Checklist or Young Adult Behavior Checklist, and self-perception was measured by the Harter Self-Perception Profile. Mean (sem) birth length sd score was -3.6 (0.2), mean age and height at start was 7.4 (0.2) yr and -3.0 (0.1) sd score, respectively, mean duration of GH treatment was 8.0 (0.2) yr, and mean age in 2001 was 16.5 (0.3) yr. After 2 yr of GH treatment, 96% of both GH groups showed a height gain sd score of 1 sd from the start of treatment or more, resulting in a normal height (i.e. height >/= -2.0 sd for age and sex) in 70% of the children. In 2001, 48 (91%) of the 53 children participating in this study had reached a normal height. Block-design s-score and the estimated total IQ significantly increased (P < 0.001 and P < 0.001, respectively) from scores significantly lower than Dutch peers at start (P < 0.001 and P < 0.001, respectively) to comparable scores in 2001. The increase over time for the Vocabulary s-score was not significant. Internalizing Behavior sd scores remained comparable to Dutch peers, whereas Externalizing Behavior sd scores and Total Problem Behavior sd scores improved significantly during GH therapy (P < 0.01 and P < 0.05, respectively) to scores comparable to Dutch peers. Self-perception sd scores improved from start of GH treatment until 2001 (P < 0.001) to scores significantly higher than Dutch peers (P < 0.05). No significant differences between the two GH dosage groups were found. Improvement in Externalizing and Total Problem Behavior sd scores over time was significantly related to change in height sd score (P < 0.05 and P < 0.01, respectively), whereas scores over time for Vocabulary, Block-design, Internalizing, or total Harter Self-Perception score were not related to change in height sd scores. In conclusion, parallel to a GH-induced catch-up growth in adolescents born SGA, IQ, behavior, and self-perception showed a significant improvement over time from scores below average to scores comparable to Dutch peers. In addition, children whose height over time became closer to that of their peers showed less problem behavior.