Pulse oximetry provides a noninvasive, painless, accurate, and rapid method for measuring arterial oxygen saturation (SaO2). It has been shown to be valuable in anesthesia and critical care and recently has been used extensively in the outpatient setting. As is often the case with new technologies, little has been published on the basic issues of reliability, reproducibility, and effect on patient care. This prospective clinical study evaluated the basic principles of pulse oximetry in the pediatric emergency department setting and tested the hypotheses that pulse oximetry SaO2 measurements are reliable and provide valuable information, in addition to clinical and laboratory data which affect patient assessment and management. One hundred twenty patients were enrolled in phase 1 and 437 in phase 2 of the study. Pulse oximetry readings were reproducible with an intraclass correlation of 0.87. SaO2 measurements changed the assessed degree of illness in 188 (53%) patients; 47 (13%) were felt to be more ill and 130 (37%) to be less ill than at initial assessment. Sixty-nine (17%) patients were identified in whom SaO2 readings changed management plans; 27 (8%) were managed more aggressively (intubation, surgery, or admission), while 40 (11%) were managed less aggressively (discharged). In three cases, pulse oximetry was instrumental in the diagnosis of a serious illness. The results from this study indicate that pulse oximetry SaO2 readings are stable and reproducible and provide information which impacts significantly on patient assessment and management.