SummaryPulse oximetry holds the promise of wide application for monitoring and assessing pulmonary function in small animal patients. Although the saturation as read by pulse oximetry (SpO2) has previously been shown to be accurate in healthy dogs, its accuracy and usefulness have not been demonstrated in critical small animal patients. The present study assessed the accuracy and usefulness of a pulse oximeter (Ohmeda Biox 3740, Ohmeda, Louisville, CO) in a small animal intensive care unit. The instrument yielded readings in 48 of 51 attempts in 33 animals (25 dogs, 8 cats). Criteria were developed to reject spurious readings; when these criteria were applied, the actual calculated SaO2 differed from the SpO2 by O.26 +2.2%, with a correlation of 0.87 (p<0.0001). The 95% confidence interval was +4.4%, comparable to the accepted level in humans. No ill effects from SpO2 were apparent in the patients, and the instrument was useful in monitoring the progress of critical animals. However, uncritical use of the oximeter could have led to gross patient mismanagement, as SpO2 readings as much as 29% different from SaO2 were sometimes obtained.