The relationship between percent of ideal dose and disease-free survival was examined in 256 Stage II and III patients who participated in a 2-year breast adjuvant chemotherapy trial consisting of cyclophosphamide, methotrexate, and 5-fluorouracil (CMF) given postoperatively. When analyzed analogously to previous work, the results confirmed a dose-response relationship: that is, there appeared to be an improved disease-free survival for patients receiving higher doses of adjuvant chemotherapy. The major criticism of such an analysis is its bias. This bias was addressed by considering only patients who were still receiving therapy at 6, 12, and 24 months; then, the dose-response relationship was no longer seen. Although causality cannot be inferred, the apparent differences in disease-free survival among the dose groups can be attributed to recurrences in the first 2 years among patients receiving lower doses of chemotherapy.