PURPOSE: We conducted a randomized Phase II trial to directly compare toxicity, feasibility, and delivered dose intensities of two adjuvant dose-intensive regimens containing doxorubicin, paclitaxel, and cyclophosphamide for patients with node-positive breast carcinoma. EXPERIMENTAL DESIGN: Forty-two patients with resected breast carcinoma involving one or more ipsilateral axillary lymph nodes, were randomized to receive two different schedules of adjuvant chemotherapy using 14-day dosing intervals: either (a) three cycles of doxorubicin 80 mg/m(2) as i.v. bolus followed sequentially by three cycles of paclitaxel 200 mg/m(2) as a 24-h infusion and then by three cycles of cyclophosphamide 3.0 g/m(2) as a 1-h infusion (arm A); or (b) the same schedule of doxorubicin followed by three cycles of concurrent cyclophosphamide and paclitaxel at the same doses (arm B). All cycles were supported by granulocyte colony-stimulating factor administration. RESULTS: Forty-one patients were assessable for toxicity and feasibility; 37 (90%) completed all planned chemotherapy. There was no treatment-related mortality; however, increased toxicity was observed on arm B compared with arm A, manifested by an increase in hospitalization for toxicity, mainly neutropenic fever, and an increased incidence of transfusion of packed RBCs transfusions for anemia. The mean delivered dose intensities for paclitaxel and cyclophosphamide were significantly greater for arm A compared with arm B (P =.01 and P =.05, respectively). There is no long-term, treatment-related toxicity, and no cases of acute myelogenous leukemia or myelodysplastic syndrome have been observed. CONCLUSIONS: Dose-dense sequential single-agent chemotherapy is more feasible than doxorubicin with subsequent concurrent paclitaxel and cyclophosphamide.