Three hundred thirty-six patients with stage II or stage III breast cancer were treated on an adjuvant protocol containing fluorouracil, doxorubicin, cyclophosphamide, vincristine, and prednisone (FACVP). Depending on the estrogen-receptor (ER) status, the patients were subdivided to receive maintenance chemotherapy with or without tamoxifen. The administered dose intensity of fluorouracil, doxorubicin, and cyclophosphamide (FAC) (mg/m2/wk) relative to the projected dose intensity (based on planned dose) was computed for each patient. The relative dose intensity of the first six cycles of chemotherapy (RDI6) was compared with disease-free survival (DFS). Of the 299 patients who completed at least six cycles of therapy, 83% received dose intensities within 20% of standard intensity (.8 less than or equal to RDI6 less than or equal to 1.2). The group with the highest dose intensity (RDI6 greater than or equal to 1.13) had the longest DFS, though there was not a clear trend of linear association between dose intensity and DFS after adjustment for prognostic factors (P = .16). The patients who received at least standard dose intensity (RDI6 greater than or equal to 1.0) had longer DFS than those whose therapy did not reach standard intensity (RDI6 less than 1.0). This difference was significant in patients with stage III disease (P = .01). The 37 patients who completed fewer than six cycles of chemotherapy had the shortest DFS (5-year DFS of 48% v 65% in the others). This retrospective analysis, in a heterogeneously treated group of patients, did not show the differences in outcome associated with dose intensity as demonstrated in the earlier studies comparing projected dose intensity of various cyclophosphamide, methotrexate-, and fluorouracil (CMF)-containing adjuvant trials. Improved DFS was noted in the stage III patients receiving higher dose intensity. Our failure to demonstrate the differences in stage II patients may be due to the narrow range of dose intensity in this study or to a difference in the dose-response curves depending on the stage of disease.