In summary, our findings support and confirm the concerns of many investigators that present methods of cardiorespiratory monitoring are inadequate for the detection of many forms of apnea. Nurses underrecord both the frequency and duration of apneic episodes. Bradycardia is an unreliable index of hypoxemia. Thoracic impedance monitors are unreliable because they detect only a fixed duration of respiratory pause and are sensitive to many artifacts unavoidable in a clinical setting. Finally, ineffective breathing patterns such as disorganized breathing, obstructive apnea, and paradoxical breathing are undetectable by thoracic impedance monitoring. We warn against the reliance on heart rate and thoracic impedance monitoring alone for infants with recurrent apnea.