Outcome from coma caused by cerebral hypoxia-ischemia (eg, cardiac arrest) was compared with serial neurological findings in 210 patients. Thirteen percent of patients regained independent function at some point during the first postarrest year. Computer application of new multivariate techniques to the prospectively observed findings generated easily utilized rules that classified patients by likely outcome. At the time of initial examination, 52 patients (one fourth of the total population) had absent pupillary light reflexes, and none of these patients ever regained independent daily function. By contrast, the initial presence of pupillary light reflexes, the development of spontaneous eye movements that were roving conjugate or better, and the findings of extensor, flexor, or withdrawal responses to pain identified a smaller group of 27 patients, 11 (41%) of whom regained independence in their daily lives. By 24 hours after onset, 93 poor-outcome patients were identified by motor responses that were absent, extensor, or flexor and by spontaneous eye movements that were neither orienting nor roving conjugate; only one regained independent function. This contrasts with recovery in 19 (63%) of 30 patients who at that time showed improvement in their eye-opening responses and obeyed commands or had motor responses that were withdrawal or localizing. Similarly simple rules distinguished between good- and poor-prognosis patients on postarrest days 3, 7, and 14.