Aims: Anthracycline-based chemotherapy improves survival in breast cancer patients but is associated with increased risk of heart failure (HF). However, the risk of late-onset HF is debatable and mainly based on observational studies. The aim of this study was to evaluate the effect of anthracycline-based chemotherapy on long-term risk of clinical HF. Methods and results: Between 1990 and 1998 the Danish Breast Cancer Cooperative Group (DBCG) 89D trial randomized 980 Danish women with early breast cancer to adjuvant cyclophosphamide, epirubicin, and fluorouracil or cyclophosphamide, methotrexate, and fluorouracil. Incident HF was the primary endpoint obtained from Danish administrative registries. Follow-up ended at December 2014. The risk of HF was evaluated in a cumulative incidence analysis and a Fine–Gray proportional hazards model. Median follow-up time was 16.9 years [interquartile range (IQR) 3.7–20.9]. In the epirubicin treatment group, 23 new cases of HF were identified vs. 9 in the non-epirubicin group corresponding to incidence rates per 1000 patient-years of 3.7 [95% confidence interval (CI) 2.5–5.6] vs. 1.4 (95% CI 0.7–2.7). The cumulative incidence of HF was higher in the epirubicin treatment group compared with the non-epirubicin group (P < 0.01), yielding a hazard ratio of 3.00 (95% CI 1.39–6.49, P < 0.01) for HF associated with epirubicin. The median dose of epirubicin was 451.9 mg/m² (IQR 400.0–523.5). Conclusions: Anthracycline-based chemotherapy is associated with a three-fold increased risk of late-onset clinical HF relative to non-anthracycline chemotherapy in this randomized clinical trial, but overall risk is low.