The Eastern Cooperative Oncology Group (ECOG) performed a prospectively randomized study (E6484) evaluating the use of interferon alfa 2a (IFN-alpha2a) in patients with aggressive low-grade or with intermediate-grade non-Hodgkin's lymphoma (NHL) accruing close to 300 patients between 1985 and 1988. Patients were eligible for study if they had bulky or symptomatic low-grade lymphoma or defined intermediate-grade subtypes. Of 291 patients enrolled, 249 were eligible for analysis. All patients were randomized to receive a four-drug cytotoxic chemotherapy regimen including cyclophosphamide, doxorubicin, vincristine and prednisone in 4-week cycles with or without IFN-alpha2a in addition (COPA vs I-COPA). Treatment was given for up to 8-10 months. This report, at a time when the median follow-up among survivors has reached 12 years, updates the analysis of time to treatment failure (TTF), duration of disease-free survival (DFS), and overall survival. Patients randomized to receive IFN-alpha2a had a prolonged TTF (P= 0.008; median 2.4 vs 1.6 years). DFS for those patients who had complete responses was also longer if IFN-alpha2a had been given (P = 0.035; median 2.7 vs 1.8 years). There was a clinically but not a statistically significant prolongation of overall survival by IFN-alpha2a (P= 0.107; median 7.8 vs 5.7 years). There were fewer deaths over time due to lymphoma in patients receiving IFN-alpha2a (67 vs 80 deaths). A subset analysis, based on disease histology (low-grade, follicular, intermediate-grade), revealed a significant prolongation of TTF in patients receiving IFN-alpha2a with either low-grade (P = 0.002; median 2.4 vs 1.6 years) or follicular (P= 0.01; median 2.5 vs 1.7 years) NHL but not intermediate grade (P = 0.622; median 2.3 vs 1.6 years) NHL. This analysis, performed approximately 12 years after closure of the study to accrual, supports the addition of interferon alfa to an induction cytotoxic chemotherapy regimen including cyclophosphamide and doxorubicin in the treatment of follicular NHL.