Reduced cardiac efficiency caused by suboptimal synchronization of the heart's normal contraction might contribute to the development of or exacerbate heart failure. Conceptually and in practice cardiac dyssynchrony is complex. Recent studies have shown that atrio-biventricular pacing can improve cardiac synchrony in many patients and improve cardiac function, symptoms and exercise capacity, and reduce morbidity and mortality substantially. Randomized controlled trials, however, indicate that the severity of cardiac dyssynchrony, as conventionally measured, is a poor guide to treatment benefit and that correction of dyssynchrony accounts for only part of the benefit of atrio-biventricular pacing. Although some of the benefits of atrio-biventricular pacing might be mediated by cardiac resynchronization, much of the benefit could be mediated by mechanisms that are as yet unknown. Withholding atrio-biventricular pacing in patients who do not exhibit cardiac dyssynchrony on imaging but otherwise fulfil the entry criteria used in randomized controlled trials of this therapy could be unwise. Here, we examine the evidence that cardiac resynchronization is indeed the mechanism by which atrio-biventricular pacing exerts its effects.