For 30 years evidence has pointed to sleep apnoea being a cause of cardiovascular (CV) disease and major CV events. However, five recent large randomised controlled trials (RCTs) of positive airways pressure treatment of sleep apnoea in high CV risk populations have reported neutral results, and in one case, harm. I review the various strands of evidence in an attempt to reconcile these conflicting results arguing that: 1) Cohort and clinical case control studies have likely overestimated the CV risk associated with sleep apnoea; 2) Rodent models have similarly overstated the risk of CV injury from sleep apnoea by focussing on the effects of extremely severe intermittent hypoxia; and 3) Significant heterogeneity in susceptibility to vascular injury may exist between patients. These factors, coupled with low adherence to positive airway pressure mask treatment, may help explain the neutral results of recent RCTs. Suggestions for future research are provided.