The randomized multicenter study on rapidly proliferating breast cancer, assessed according to thymidine labelling index (TLI), was activated at the end of the 1980s. The present work investigated whether and to what degree the short-term advantages observed from adjuvant CMF (cyclophosphamide, methotrexate, 5-fluorouracil) were maintained at a longer follow-up. Two hundred and eighty-one patients with node-negative and high TLI tumors were randomized to receive six cycles of CMF or no further treatment. At a median follow-up of 12 years, CMF produced a 25% and 20% relative reduction in relapse and death cumulative incidence, respectively. A breakdown analysis identified a subgroup of patients with intermediate proliferating tumors for whom a 70% and 73% reduction in relapse and death was observed in the intention-to-treat population. An even higher reduction of 80% and 84% in relapse and death was seen for the patients who had received the full CMF dose. We identified a subgroup of patients with intermediate proliferating tumors in whom the high benefit obtained from adjuvant CMF was maintained at a long-term follow up.