Using a 2 x 2 factorial design, we studied the adjuvant chemotherapy of women with axillary node-positive breast cancer to compare sequential doxorubicin (A), paclitaxel (T), and cyclophosphamide (C) with concurrent doxorubicin and cyclophosphamide (AC) followed by paclitaxel (T) for disease-free (DFS) and overall survival (OS); to determine whether the dose density of the agents improves DFS and OS; and to compare toxicities. A total of 2,005 female patients were randomly assigned to receive one of the following regimens: (I) sequential A x 4 (doses) --> T x 4 --> C x 4 with doses every 3 weeks, (II) sequential A x 4 --> T x 4 --> C x 4 every 2 weeks with filgrastim, (III) concurrent AC x 4 --> T x 4 every 3 weeks, or (IV) concurrent AC x 4 --> T x 4 every 2 weeks with filgrastim. A protocol-specified analysis was performed at a median follow-up of 36 months: 315 patients had experienced relapse or died, compared with 515 expected treatment failures. Dose-dense treatment improved the primary end point, DFS (risk ratio RR = 0.74; P =.010), and OS (RR = 0.69; P =.013). Four-year DFS was 82% for the dose-dense regimens and 75% for the others. There was no difference in either DFS or OS between the concurrent and sequential schedules. There was no interaction between density and sequence. Severe neutropenia was less frequent in patients who received the dose-dense regimens. Dose density improves clinical outcomes significantly, despite the lower than expected number of events at this time. Sequential chemotherapy is as effective as concurrent chemotherapy.