Background: Previous evaluations of survivors of childhood cancer suggest that the cumulative incidence of secondary leukemia plateaus at ~2% 10-15 years after primary cancer therapy. Risk beyond 15 years has not been comprehensively assessed, in part because of the lack of comprehensive long-term follow-up of previous cohorts beyond this time period. Methods: The Childhood Cancer Survivor Study (CCSS) is a retrospective cohort study of 5-year survivors of childhood cancer treated between 1970 and 1986. Self-reported secondary leukemias >5 years from primary diagnosis were confirmed by pathology report or medical records. Diagnostic bone marrow samples were acquired and centrally reviewed. The standardized incidence ratio (SIR) was derived using age, gender, and calendar year specific rates from surveillance epidemiology and end results data. Results: Of 14,358 5-year survivors, 43 developed subsequent leukemia >5 years from primary diagnosis with 13 cases of leukemia occurring ≥ 15 years from diagnosis. Five were AML (2 with confirmed preceding MDS), 4 ALL (2 pre-B lineage, 1 T-cell, 1 unknown), 2 were APL with t(15;17). Two AML cases had 7q- deletion. One ALL had a complex karyotype including t(9;22) and another a p53 mutation. The mean age at diagnosis of leukemia was 31.6 years (range 18 to51), and occurred after a latency of 21.6 years from diagnosis. Sarcomas (n=5) and Hodgkin lymphoma (n=4) were the most common primary diagnoses. Six of the 13 patients received radiation therapy, while only 5 received alkylating agents, and none received epipodophyllotoxins. The SIR for leukemias occurring ≥ 15 years from diagnosis was 3.5 (95% CI 1.9 - 6.0). Eight of 13 are deceased with a median survival of 2 years after diagnosis of leukemia. Conclusions: This is the first description of increased risk of secondary leukemia ≥ 15 years from primary malignancy, demonstrating a 3.5-fold increased risk above that of the general population. Radiation was the most common treatment exposure. These findings indicate a high level of suspicion should be maintained for subjects who received radiation or alkylators and develop pancytopenia to facilitate early detection and intervention.