The European Cooperative Trial in Operable breast cancer (ECTO) randomly tested whether efficacy of adjuvant doxorubicin followed by i.v. cyclophosphamide, methotrexate, and fluorouracil (CMF; doxorubicin-->CMF, arm A) could be improved by adding paclitaxel (doxorubicin/paclitaxel-->CMF) as adjuvant (arm B) or primary systemic therapy (PST, arm C). We report here feasibility, tolerability, locoregional antitumor activity, and breast conservation rate. A total of 1,355 women entered the study. Feasibility and safety were compared in arm A versus arms B plus C. Surgical findings were compared in arms A plus B versus arm C. Grade 3 or 4 National Cancer Institute toxicities were low (<5%) in all arms. Neuropathy was more frequent in the paclitaxel-containing arms (grade 2, 20.5% versus 5.0%; grade 3, 1.3% versus 0.2%). At 31 months of follow-up, asymptomatic drop of left ventricular ejection fraction was similar in all arms, whereas symptomatic cardiotoxicity was recorded in three patients (0.5%) in A and in three patients (0.3%) in B plus C. PST induced clinical complete plus partial remission in 78%, with an in-breast pathologic complete response rate of 23% and an in-breast plus axilla pathologic complete response rate of 20%. In the multivariate analysis, only estrogen receptor (ER) status was significantly associated with pathologic complete response (odds ratio for ER negative, 5.77; 95% confidence interval, 3.49-9.52; P<0.0001). PTS induced a significant axillary downstaging (P<0.001), and breast sparing surgery was feasible in 65% versus 34% (P<0.001). Doxorubicin/paclitaxel-->CMF is feasible, safe, and well tolerated. Given as PST, it is markedly active, allowing for breast-sparing surgery in a large fraction of patients.