Aims: Current guidelines recommend implantable cardioverter-defibrillator (ICD) therapy in patients with heart failure, a left ventricular ejection fraction of ≤35%, and New York Heart Association (NYHA) class II–III. However, the evidence regarding the benefit of primary prevention ICD is less consistent in patients with NYHA class III. We investigated the long-term effects of primary prevention ICD implantation according to NYHA class in an extended follow-up study of the DANISH trial. Methods and results: The DANISH trial randomized 1116 patients with non-ischaemic heart failure with reduced ejection fraction (HFrEF) to ICD implantation or usual care. Outcomes were analysed according to NYHA class at baseline (NYHA class II and III/IV). The primary outcome was all-cause mortality. Of the 1116 patients randomized in the DANISH trial, 597 (53.5%) were in NYHA class II at baseline, 505 (45.3%) in NYHA class III, and 14 (1.3%) in NYHA class IV. During a median follow-up of 9.5 years, NYHA class III/IV, compared with NYHA class II, were associated with a greater long-term rate of all-cause mortality (hazard ratio [HR] 1.52, 95% confidence interval [CI] 1.20–1.93) and cardiovascular death (HR 1.95 [1.47–2.60]). ICD implantation, compared with usual care, did not reduce the long-term rate of all-cause mortality (all participants: HR 0.89 [95% CI 0.74–1.08]; NYHA class II: HR 0.85 [0.64–1.13]; NYHA class III/IV: HR 0.89 [0.69–1.14]; p_interaction = 0.78) or cardiovascular death (all participants: HR 0.87 [95% CI 0.70–1.09]; NYHA class II: HR 0.78 [0.54–1.12]; NYHA class III/IV: HR 0.89 [0.67–1.19]; p_interaction = 0.58), irrespective of NYHA class. Similarly, NYHA class did not modify the beneficial effects of ICD implantation on sudden cardiovascular death (all participants: HR 0.60 [95% CI 0.40–0.92]; NYHA class II: HR 0.73 [0.40–1.36]; NYHA class III/IV: HR 0.52 [0.29–0.94]; p_interaction = 0.39). Conclusions: In patients with non-ischaemic HFrEF, ICD implantation, compared with usual care, did not reduce the overall mortality rate, but it did reduce sudden cardiovascular death, regardless of baseline NYHA class. Clinical Trial Registration: ClinicalTrials.gov NCT00542945.