Assessment of the accuracy of non-invasive rebound tonometry, and comparison with invasive cannulation tonometry. An in vivo calibration technique was devised to improve the accuracy of the rebound tonometer. IOP was then measured in SW mice using both rebound and cannulation tonometry. The ability of the rebound tonometer to accurately measure small IOP reductions after instillation of a topical prostaglandin was also determined. With the rebound method, mid-afternoon IOP in two groups of similar aged SW mice was 15.9+/-3.9 mmHg (mean+/-s.d., n=25) compared to 16.3+/-1.2 mmHg (n=32) using the cannulation technique. This difference was not statistically significant (p=0.6). For serial measurements using both techniques in the same eyes of a third group of SW mice (n=14), mean IOP was 15.0+/-3.9 mmHg for rebound tonometry but only 13.4+/-2.3 mmHg for subsequent cannulation tonometry. This effect was subsequently shown to be a consequence of the rebound tonometry, as multiple rebound measurements induced a statistically significant reduction in IOP. The average IOP reduction observed 2 hr after a single application of topical latanoprost (200 ng) was 2.8+/-1.3 mmHg (p<0.001) and 2.4+/-4.7 mmHg (p=0.03) with cannulation and rebound tonometers, respectively. These differences were not significantly different (p=0.8). In vivo calibration of the rebound tonometer increased measurement accuracy and provided IOP values within the physiological range that agreed closely with the IOP measured by cannulation tonometry. However, IOP measurement with the rebound tonometer had larger variability compared with the cannulation method. Repeat IOP measurements with the rebound tonometer led to a reduction in IOP. The rebound tonometer was sufficiently sensitive to detect a 2-3 mmHg reduction in IOP following application of topical latanoprost. Despite these limitations, the rebound tonometer has a significant advantage over cannulation tonometry in that it permits longitudinal IOP measurement in conscious mice.