To determine if temperature regulation is improved during neonatal transport using a servo-regulated cooling device when compared with standard practice. We performed a multicenter, randomized, nonmasked clinical trial in newborns with neonatal encephalopathy cooled during transport to 9 neonatal intensive care units in California. Newborns who met institutional criteria for therapeutic hypothermia were randomly assigned to receive cooling according to usual center practices vs device servo-regulated cooling. The primary outcome was the percentage of temperatures in target range (33°-34°C) during transport. Secondary outcomes included percentage of newborns reaching target temperature any time during transport, time to target temperature, and percentage of newborns in target range 1 hour after cooling initiation. One hundred newborns were enrolled: 49 to control arm and 51 to device arm. Baseline demographics did not differ with the exception of cord pH. For each subject, the percentage of temperatures in the target range was calculated. Infants cooled using the device had a higher percentage of temperatures in target range compared with control infants (median 73% [IQR 17-88] vs 0% [IQR 0-52], P < .001). More subjects reached target temperature during transport using the servo-regulated device (80% vs 49%, P <.001), and in a shorter time period (44 ± 31 minutes vs 63 ± 37 minutes, P = .04). Device-cooled infants reached target temperature by 1 hour with greater frequency than control infants (71% vs 20%, P < .001). Cooling using a servo-regulated device provides more predictable temperature management during neonatal transport than does usual care for outborn newborns with neonatal encephalopathy.