To evaluate the importance of pretreatment karyotype in predicting long-term outcome in acute lymphoblastic leukemia (ALL), we performed a follow-up study of the 329 patients from the Third International Workshop on Chromosomes in Leukemia. Living patients have now been followed a minimum of 6 years. Patients were divided into ten groups according to pretreatment karyotype: no abnormalities, one of the following structural abnormalities [the Philadelphia chromosome, rearrangements involving 8q24, t(4;11), 14q+, 6q-] or, in the remaining cases, modal number (less than 46, 46, 47-50, greater than 50). As previously reported for achievement and duration of complete remission, and overall survival, disease-free survival differed significantly (p less than 0.001) among chromosome groups for both adults and children. Among children, karyotype was an independent prognostic factor for predicting disease-free survival. Because of the long follow-up, we now have been able to utilize statistical models to estimate the percentage of patients cured, according to karyotype alone and combined with other risk factors. Adults with the highest likelihood of cure (21-33%) were those patients with FAB-L1, a leukocyte count of 50,000/microliters or less, and one of the following chromosome groups: greater than 50, 47-50, 6q-, or normal. In children these same characteristics were associated with the highest percentage of cure (58-71% cured). In addition, we identified several groups of children with less than 15% chance of cure who clearly need to be treated as high-risk patients at diagnosis. Future studies of patients who have received risk-adapted therapy based on these chromosome data are needed to determine if more intensive treatment will improve the outlook of patients with cytogenetically unfavorable types of ALL.