The CArdiac REsynchronization-Heart Failure study randomized patients with left ventricular ejection fraction < or =35%, markers of cardiac dyssynchrony, and persistent moderate or severe symptoms of heart failure despite pharmacological therapy, to implantation of a cardiac resynchronization therapy (CRT) device or not. The main study observed substantial benefits on morbidity and mortality during a mean follow-up of 29.4 months [median 29.6, interquartile range (IQR) 23.6-34.6]. Prior to study closure, an extension phase lasting a further 8 months (allowing time for data analysis and presentation) was declared during which cross-over was discouraged. This was an extension of the already reported open-label randomized trial described above. The primary outcome of the extension phase was all-cause mortality from the time of randomization to completion of the extension phase. The secondary outcome was mode of death. The mean follow-up was 37.4 months (median 37.6, IQR 31.5-42.5, range 26.1-52.6 months). There were 154 deaths (38.1%) in 404 patients assigned to medical therapy and 101 deaths (24.7%) in 409 patients assigned to CRT (hazard ratio 0.60, 95% CI 0.47-0.77, P<0.0001) without evidence of heterogeneity in pre-specified subgroups. A reduction in the risk of death due to heart failure (64 vs. 38 deaths; hazard ratio 0.55, 95% CI 0.37-0.82, P=0.003) and sudden death was observed (55 vs. 32; hazard ratio 0.54, 95% CI 0.35-0.84, P=0.005). The benefits of CRT observed in the main trial persist or increase with longer follow-up. Reduction in mortality was due to fewer deaths both from worsening heart failure and from sudden death.