To prospectively validate the use of a simplified geriatric assessment (sGA) at diagnosis and to integrate it into a prognostic score for older patients with diffuse large B-cell lymphoma (DLBCL).We conducted the prospective Elderly Project study on patients with DLBCL older than 64 years who underwent our Fondazione Italiana Linfomi original geriatric assessment (oGA) (age, Cumulative Illness Rating Scale for Geriatrics, activities of daily living, and instrumental activities of daily living) before treatment. Treatment choice was left to the physician's discretion. The primary end point was overall survival (OS) (ClinicalTrials.gov identifier: NCT02364050).We analyzed 1,163 patients (median age 76 years), with a 3-year OS of 65% (95% CI, 62 to 68). Because at multivariate analysis on oGA, age > 80 years retained an independent correlation with OS, we also developed a new, simplified version of the GA (sGA) that classifies patients as fit (55%), unfit (28%), and frail (18%) with significantly different 3-year OS of 75%, 58%, and 43%, respectively. The sGA groups, International Prognostic Index, and hemoglobin levels were independent predictors of OS and were used to build the Elderly Prognostic Index (EPI). Three risk groups were identified: low (23%), intermediate (48%), and high (29%), with an estimated 3-year OS of 87% (95% CI, 81 to 91), 69% (95% CI, 63 to 73), and 42% (95% CI, 36 to 49), respectively. The EPI was validated using an independent external series of 328 cases.The Elderly Project validates sGA as an objective tool to assess fitness status and defines the new EPI to predict OS of older patients with DLBCL.