Importance: Trials of adjuvant high-dose chemotherapy (HDCT) have failed to show a survival benefit in unselected patients with breast cancer, but long-term follow-up is lacking. Objective: To determine 20-year efficacy and safety outcomes of a large trial of adjuvant HDCT vs conventional-dose chemotherapy (CDCT) for patients with stage III breast cancer. Design, Setting, and Participants: This secondary analysis used data from a randomized phase 3 multicenter clinical trial of 885 women younger than 56 years with breast cancer and 4 or more involved axillary lymph nodes conducted from August 1, 1993, to July 31, 1999. Additional follow-up data were collected between June 1, 2016, and December 31, 2017, from medical records, general practitioners, the Dutch national statistical office, and nationwide cancer registries. Analysis was performed on an intention-to-treat basis. Statistical analysis was performed from February 1, 2018, to October 14, 2019. Interventions: Participants were randomized 1:1 to receive 5 cycles of CDCT consisting of fluorouracil, 500 mg/m 2, epirubicin, 90 mg/m 2, and cyclophosphamide, 500 mg/m 2, or HDCT in which the first 4 cycles were identical to CDCT and the fifth cycle was replaced by cyclophosphamide, 6000 mg/m 2, thiotepa, 480 mg/m 2, and carboplatin, 1600 mg/m 2, followed by hematopoietic stem cell transplant. Main Outcomes and Measures: Main end points were overall survival and safety and cumulative incidence risk of a second malignant neoplasm or cardiovascular events. Results: Of the 885 women in the study (mean [SD] age, 44.5 [6.6] years), 442 were randomized to receive HDCT, and 443 were randomized to receive CDCT. With 20.4 years median follow-up (interquartile range, 19.2-22.0 years), the 20-year overall survival was 45.3% with HDCT and 41.5% with CDCT (hazard ratio, 0.89; 95% CI, 0.75-1.06). The absolute improvement in 20-year overall survival was 14.6% (hazard ratio, 0.72; 95% CI, 0.54-0.95) for patients with 10 or more invoved axillary lymph nodes and 15.4% (hazard ratio, 0.67; 95% CI, 0.42-1.05) for patients with triple-negative breast cancer. The cumulative incidence risk of a second malignant neoplasm at 20 years or major cardiovascular events was similar in both treatment groups (20-year cumulative incidence risk for second malignant neoplasm was 12.1% in the HDCT group vs 16.2% in the CDCT group, P =.10), although patients in the HDCT group more often had hypertension (21.7% vs 14.3%, P =.02), hypercholesterolemia (15.7% vs 10.6%, P =.04), and dysrhythmias (8.6% vs 4.6%, P =.005). Conclusions and Relevance: High-dose chemotherapy provided no long-term survival benefit in unselected patients with stage III breast cancer but did provide improved overall survival in very high-risk patients (ie, with ≥10 involved axillary lymph nodes). High-dose chemotherapy did not affect long-term risk of a second malignant neoplasm or major cardiovascular events. Trial Registration: ClinicalTrials.gov Identifier: NCT03087409.