An American College of Cardiology (ACC)/American Heart Association (AHA) task force on practice guidelines in 2001 published evidence-based recommendations for performing percutaneous coronary interventions (PCIs). These guidelines grouped the indications for PCI into 4 classes (I, IIa, IIb, and III) based on analyses of risks and benefits. In a previous study, we found that clinical success and in-hospital adverse events varied by indications class. However, no adjustment for risk was used in those comparisons. The ACC/National Cardiovascular Data Registry (ACC-NCDR) previously developed a risk-adjustment model for the adverse event of in-hospital PCI mortality. We investigated how the 14 individual risk factors in the ACC-NCDR PCI mortality model might differ across the 4 indications classes and whether estimated mortality for each class approximated the observed mortality for that class. We analyzed the ACC-NCDR PCI database for January 1, 2001 to December 31, 2004. We excluded procedures performed for treatment of acute ST-segment elevation myocardial infarction; all others were included, yielding 559,273 procedures for analysis. An algorithm derived from the 2001 guidelines was used to assign procedures to an indications class. Increasing frequencies of risk components were observed across classes I, IIa, IIb, and III. Expected mortalities for each class calculated by the risk-adjustment model were close to observed values (expected 0.52%, 0.59%, 1.72%, and 1.96%, respectively; observed 0.49%, 0.63%, 1.88%, and 1.60%, respectively). In conclusion, the ACC-NCDR risk-adjusted mortality model can be linked to the ACC/AHA PCI guidelines, and together these produce mortality risk estimates by indications classes that are close to actual observed values. With further refinement, these methods should be able to be used as powerful analytic tools for quality assurance and appropriateness purposes.