A scoring system was developed to objectively evaluate the condition of transported preterm infants. The "transport score" used five variables: blood glucose, blood pressure, pH, pO2, and temperature. Each variable was scored 0, 1, or 2, with a total achievable score from 0 to 10. Twenty-one matched pairs of infants included one infant who lived and one who died. The transport scores upon admission of infants who lived was significantly greater than the scores of those who died (P less than 0.01). Scores less than eight were predictive of death (sensitivity 62%, specificity 81%). The system was then applied prospectively to 106 different infants after stabilization by the hospital-of-origin (pre-transport) and upon admission to the neonatal intensive care unit (post-transport). Although 75 (76%) of the 99 surviving infants had both stabilization and admissions scores of eight or more, only 2 (29%) of the 7 infants who died had both scores of eight or more. Of 85 infants with a stabilization score of eight or greater, only 3 (3.5%) died, while of 21 infants with stabilization scores less than eight, 4 (19%) died. Similarly, of 90 infants with an admission score of eight or more, only 4 (4.4%) died, while of 16 with an admission score of less than eight, 3 (19%) died. We conclude that transport scores provide a valid indication of the condition of preterm infants and may be used to provide quality assurance for stabilization and transport efforts.