New-onset post-coronary artery bypass graft surgery (CABG) atrial fibrillation (AF) is associated with poor outcomes, but data on the effects of its characteristics are lacking and conflicting. We examined the effect number of post-CABG AF events has on long-term mortality risk, and whether this is sex dependent.Routinely-collected Society of Thoracic Surgeons (STS) data were supplemented with details on new-onset post-CABG AF (detected in-hospital via continuous ECG/telemetry monitoring) and long-term survival for 9,203 consecutive isolated-CABG patients (2002-2010). Using Cox regression, we determined the propensity-adjusted (STS-recognized risk factors) effect of number of AF events on survival, testing for effect modification by sex and controlling for AF duration.AF occurred in 739 (29.4%) women and 2,157 (32.3%) men (p<0.001). Adjusted results showed ≥2 AF events significantly (p<.0001) increased 5-year mortality risk, independently of total AF duration. However, mortality risk differed between the sexes (p<.0001): women with 2 AF episodes had the greatest increase (hazard ratio [HR]; 95% confidence interval [CI]: 2.98; 1.43,4.83, vs. women without AF), followed by women and men with ≥4 AF events (HR; 95%CI: 2.76; 1.27,4.55 and 2.73; 2.30,3.19, respectively). A single post-CABG AF episode was not associated with increased mortality risk.Both men and women who experienced ≥2 post-CABG AF episodes showed increased risk of five-year mortality, independent of total AF duration. While men's risk increased as number of AF events increased, women's peaked at 2 AF events. Future research needs to determine whether this divergence stems from differences in treatment/management, or underlying biology.