Objective: Young adults with childhood-onset GH deficiency (GHD) have reduced memory and attention, which can be improved by treatment with GH. Little information is available on cognitive function in elderly GHD patients. Design: Single center, double-blind, randomized, placebo-controlled study of 52-week duration. Methods: Elderly GH therapy naïve GHD patients (n = 34; age range 60-77 years) were enrolled and randomized to receive placebo or GH therapy which was titrated to achieve a target IGF-I level of + 1 to + 2 S.D. of the normal mean for age. Cognitive function was assessed at baseline and after 2 4 and 52 weeks, using a computerized psychometric test package (Neurobehavioral Examination System-2). Results: The mean GH dose was 0. 16 ± 0.06 mg/day; mean IGF-I increased from 135 ± 59 ng/ml at baseline to 213 ± 77 ng/ml during active treatment. The GH-treated group had better mean serial digit learning scores compared with placebo group (P<0.05). Assessment of effect sizes showed that improvements in memory occurred with GH after 24 weeks. The overall adverse event rates were similar in the GH and the placebo group. Conclusion: This study indicates that GH replacement may be accompanied by improvement in certain measures of cognitive function in elderly patients with GHD.