The American College of Cardiology/American Heart Association (ACC/AHA) and the American College of Physicians (ACP) have disseminated guidelines to assess preoperative cardiac risks before noncardiac surgery. The objectives of this study were to determine if these guidelines differ in preoperative recommendations for a group of patients, and whether these recommendations differ from actual provider recommendations. In this retrospective cohort study, patient characteristics and physician recommendations were abstracted from electronic medical records of consecutive patients attending a Veteran Affairs medical preoperative evaluation clinic from January 1 to April 1, 1998. Patient characteristics were used to determine what preoperative cardiac testing should have been ordered if each guideline was followed. Possible recommendations included operation without testing (OWT), noninvasive stress testing (NST), cardiac catheterization (CC), or cancel or delay surgery (OTHER). Recommendations were compared using statistical tests for agreement. Of the 138 patients identified, most underwent moderate-risk surgeries. Recommendations for preoperative testing were discordant between guidelines for 17% of patients (kappa = 0.38). Guidelines never agreed on the need for NST. Extreme differences in recommendations (i.e., one recommends OWT, the other CC) occurred in 9 patients (7%). Physicians ordered NST more often (n = 27) than either guideline. In this subgroup of patients where providers ordered a NST, the 2 guidelines significantly differed (kappa = 0.26). When applied to real patients being evaluated for surgery, ACC/AHA and ACP guidelines significantly differed in recommendations for preoperative cardiac testing. Results have implications for implementation, management, and practitioner adherence to published guidelines.