GH deficiency is common in survivors of childhood cancer, especially in those treated with radiation to the brain. The impact of GH therapy on disease recurrence has been studied in survivors of pediatric brain tumors, but few data are available on the risk of disease recurrence in survivors of other tumor types who are treated with GH. Likewise, the risk of second neoplasms (SN) associated with GH use has not been systematically evaluated. We studied 361 GH-treated cancer survivors (including 172 brain tumor survivors) from among 13,539 survivors enrolled in the Childhood Cancer Survivor Study, a cohort of 5-yr survivors of childhood cancer. Using a time-dependent Cox model, we compared risk of recurrence, risk of SN, and risk of death between survivors who did and did not receive treatment with GH. The relative risk of disease recurrence was 0.83 (95% confidence interval, 0.37-1.86; P = 0.65) for GH-treated survivors. The relative risk of recurrence was not increased for any of the major cancer diagnoses. GH-treated subjects were diagnosed with 15 SN, all solid tumors and no secondary leukemias, for an overall relative risk of 3.21 (95% confidence interval, 1.88-5.46; P < 0.0001). This was mainly due to a small excess number of SN observed in GH-treated survivors of acute leukemia. The risk of death was not associated with GH use (P = 0.43). We conclude that GH therapy does not appear to increase the risk of disease recurrence or death in survivors of childhood cancer. The increased number of SN, particularly in survivors of acute leukemia, is of concern, but the data need to be interpreted with caution given the small number of events.