Diagnosing lung function abnormalities requires application of the appropriate reference equation for a given patient population. Current guidelines recommend the National Health and Examination Study III data set for evaluating patients in the United States. In Caucasian patients, relying on older reference equations, as opposed to those derived from the NHANES III data set, will often result in a different interpretation of a patient's spirometry. The present study assessed whether similar discordance would occur in African-American patients. A cross-sectional analysis of African-American patients undergoing spirometry testing at our hospital was performed. Patients were classified as normal, restricted, obstructed or mixed based upon the ATS/ERS guidelines, using Crapo, Knudson, Morris, Glindmeyer, and NHANES III prediction equations. Differences in classification were evaluated. 4463 subjects were identified, with a mean age of 49.6. Discordance in interpretation was most common when results from prediction equations by Morris, Knudson, and Glindmeyer were compared to NHANES III (24.6%, 26.4%, and 20.1%, respectively). Discordance was less common when comparing Crapo to NHANES III (12.8%). There was a tendency for Knudson, Morris and Glindmeyer to under classify restriction, and for Crapo, Morris, and Glindmeyer to over classify obstruction. There is significant discordance in interpretation when spirometry for African-American patients is referenced to equations published by Crapo, Morris, Knudson, and Glindmeyer, compared to NHANES III.