Therapeutic hypothermia holds promise as a rescue neuroprotective strategy for hypoxic-ischemic injury, but the incidence of severe neurologic sequelae with hypothermia is unknown in encephalopathic neonates who present shortly after birth. This study reports a multicenter, randomized, controlled, pilot trial of moderate systemic hypothermia (33 degrees C) vs normothermia (37 degrees C) for 48 hours in neonates initiated within 6 hours of birth or hypoxic-ischemic event. The trial tested the ability to initiate systemic hypothermia in outlying hospitals and participating tertiary care centers, and determined the incidence of adverse neurologic outcomes of death and developmental scores at 12 months by Bayley II or Vineland tests between normothermic and hypothermic groups. Thirty-two hypothermic and 33 normothermic neonates were enrolled. The entry criteria selected a severely affected group of neonates, with 77% Sarnat stage III. Ten hypothermia (10/32, 31%) and 14 normothermia (14/33, 42%) patients expired. Controlling for treatment group, outborn infants were significantly more likely to die than hypoxic-ischemic infants born in participating tertiary care centers (odds ratio 10.7, 95% confidence interval 1.3-90). Severely abnormal motor scores (Psychomotor Development Index < 70) were recorded in 64% of normothermia patients and in 24% of hypothermia patients. The combined outcome of death or severe motor scores yielded fewer bad outcomes in the hypothermia group (52%) than the normothermia group (84%) (P = 0.019). Although these results need to be validated in a large clinical trial, this pilot trial provides important data for clinical trial design of hypothermia treatment in neonatal hypoxic-ischemic injury.