AIMS--The IOP variation on repeated testing with the recently introduced Keeler Pulsair 2000 instrument was investigated. METHODS--One hundred normal individuals (50 male and 50 female) new to the instrument had three sets of IOP recordings within a 15 minute time period. RESULTS--The mean of the first set of IOPs from both right and left eyes was significantly higher than those from subsequent sets (p < 0.0001 for right eyes and p = 0.01 and < 0.0002 for left eyes). This tendency increased significantly with increasing IOP. Second and third IOP sets were, however, similar indicating stabilisation of IOP measurements. The coefficient of repeatability of the instrument between second and third sets was 4.2 mm Hg for right eyes and 3.6 mm Hg for left eyes. CONCLUSION--The Pulsair 2000 passes the British standard for reproducibility of a standard test method.