The state of the art of spirometric instrumentation was discussed. Performance studies conducted in 1980 showed that eight of 12 volume type spirometers performed acceptably and that none of seven flow type spirometers performed adequately. Less than half of the instruments were computerized. Performance trials of 62 spirometers, 95% of which were computerized, conducted in 1989 showed that only 35% performed acceptably. This indicated that despite the large increase in the number of computerized spirometry systems between 1980 and 1989, this was not accompanied by a significant increase in their quality. Approximately 25% of the spirometers tested in 1989 had software problems. Some of these performed acceptably after the software was modified. This was taken as evidence that most manufacturers had not rigorously tested their spirometers before releasing them for evaluation. It was suggested that all new spirometers be comprehensively tested using the ATS 24 standard wave forms. The author concludes that although there has been an increase in the number of computerized spirometry systems over the past decade which makes them easier to use their accuracy may not have improved.