The Lymphoma Study Group (LSG) of the Japan Clinical Oncology Group conducted a phase II trial of LSG12 therapy for 45 elderly patients with aggressive lymphoma to clarify whether LSG12 reduces severe infection without lowering the complete response (CR) rate in comparison with LSG4. LSG12, which consisted of a regimen of vincristine, cyclophosphamide, prednisolone, doxorubicin, vindesine, etoposide, and procarbazine (VEPA/FEPP), excluded bleomycin and methotrexate of LSG4 therapy, reduced the dosages of doxorubicin and cyclophosphamide, and increased etoposide and procarbazine dosages instead. Inclusion criteria consisted of a patient age of 70 to 75 years, a World Health Organization performance status of 0 to 2, and acceptable organ function. The treatment was completed in 47% of the patients and terminated early for disease progression in 20% and for toxicity in 16%. The CR rate was 60% (95% confidence interval [CI], 44%-74%). The 5-year overall survival (OS) rate was 42% (95% CI, 27%-57%), and the median OS time was 4.3 years. Leukopenia of grade 3 to 4 occurred in 98% of the patients, and severe infection occurred in 9%. Eight patients with hepatitis C virus (HCV) antibody showed no severe hepatic toxicity and had a better CR or OS rate than the 37 HCV-negative patients. Although the outcomes of LSG12 met our expectations with a reduction in severe infection and equivalent CR and OS outcomes compared with LSG4 and CHOP (cyclophosphamide, doxorubicin, vincristine, and prednisolone), the possibility of a regimen more beneficial than LSG12 for aggressive lymphoma in the elderly patient should be explored because of frequent hematologic toxicity and poor compliance in LSG12.