Advanced age is an adverse prognostic factor in patients with DLCL. CHOP (cyclophosphamide, doxorubicin, vincristine, and prednisone) has frequent dose-limiting toxicities, including myelosuppression. We retrospectively reviewed 50 consecutive patients > 60 years of age (median age 72) with B-cell DLCL who received CHOP with G-CSF. Patients received CHOP (median 6 cycles) at three-week intervals. G-CSF was given following all cycles of chemotherapy ("prophylactic G-CSF") in 28 of 50 patients, and following an episode of febrile neutropenia and thereafter in 19 patients, according to ASCO guidelines. Dose intensity, treatment delays, episodes of febrile neutropenia, complete response (CR) rate, disease-free survival, time-to-treatment failure, and overall survival were all analyzed according to the age-adjusted International Prognostic Index (aaIPI). The actual dose intensity for cyclophosphamide was 225.9 mg/m2/week and 0.90, respectively and for doxorubicin was 14.9 mg/m2/week (90% of ideal CHOP dosing for both drugs). Median followup was 4 years for the patients still living. Treatment delays and episodes of febrile neutropenia were less frequent among patients receiving G-CSF with all cycles of CHOP. The CR rates were 100%, 81%, 85%, and 36% for the low, low-intermediate, high-intermediate, and high aalPI risk groups, respectively. The 5-year actuarial relapse-free and overall survival for our patients were comparable to that of the cohort < or = 60 years of age and superior to the > 60 years of age cohort used to establish the aaIPI. With optimization of CHOP dosing, advanced age may not be an adverse prognostic factor for patients with DLCL. The routine use of G-CSF in elderly patients with DLCL should be further investigated.