OBJECTIVES: Inpatient surgery is a major component of overall health care spending. We examined variation in outlier payment across US hospitals and the extent to which variation is explained by quality of care. METHODS: We used the 2006 Medicare Provider Analysis and Review (MEDPAR) file. We identified coronary artery bypass grafting (CABG).We first describe the incidence of outlier payments for CABG, average outlier payment amount and their contribution to overall inpatient payments. We then explore how outlier payments vary according to patient characteristics and across hospitals. Multiple logistic regression is used to examine to extent to which different factors serve as independent predictors of outlier payments. Standard errors are adjusted for the effect of clustering of outlier payments within hospitals. In assessing variation in outlier payments across hospitals, we described the distribution of outlier payment prevalence (proportion of patients associated with outlier payments) by hospital. We, then assessed hospital variation using fixed-efects logistic regression models. RESULTS: The proportion of patients associated with outlier payment was 11% . Average outlier payments were considerable: $26,064. Outlier payment for CABG cost CMS approximately $480 million in 2006. Outlier payments were major contributors to the overall inpatient cost: 12.9%. Approximately 20% of hospitals had outlier rates below 5% for coronary artery bypass surgery, while 25% had outlier rates exceed 20%. Although, there were patient level risk factors that determine patient level otulier payment rates, this did not explain hospital level variation. Higher volume hospitals were less likely to have patients with outlier payments. CONCLUSIONS: Aiming to accelerate the quality improvements, payers are increasingly applying value-based purchasing strategies to surgical care.We showed that the variation in outlier payments across US hospitals is substantial for CABG and patient level risk factors can not explain hospital level variation. Hospital and surgical volume as a quality indicator is negatively related with outlier payments.