Selection of spirometric test values for reporting and interpretation has recently received considerable attention. In 1977, the American Thoracic Society (ATS) Snowbird Workshop on Standardization of Spirometry recommended that the maximal values for FVC and FEV1 be used for clinical interpretation, even if they came from different spirometric tracings. The Intermountain Thoracic Society (ITS) had recommended in 1975 that FVC and FEV1, be reported from the single tracings, using the largest sum of FVC plus FEV1 (best test). We evaluated the results of 1,853 spirometric test sessions in 1,101 subjects (923 hospital patients and 178 normal volunteers). The mean difference between the 2 test selection methods cited above was 5.8 ml for FVC and 8.4 ml for FEV1. In 98.4% of the FVC comparisons and 95.7% of the FEV1 comparisons, the differences were within the minimal instrument accuracy standard (+/- 50 ml or +/- 3% of the reading) suggested by the ATS. Differences between maximal and best test FVC and FEV1 were small. The selection of values for interpretation from the best test did not compromise accuracy, and was a simpler and more practical method for reporting clinical spirometric results.