Patient care by medical school physicians is viewed as more costly compared to the private sector. Objective information concerning the cost and efficiency of neonatal care analyzed by provider group is not available. We studied factors effecting the length of hospital stay and resulting hospital charges among patients managed by two separate groups of neonatologists. Retrospectively, we analyzed the clinical and financial data base information of all neonates admitted in fiscal year 1994 to the same 100 bed, acute care nursery in a large, urban university children's hospital setting by a full time academic division and a private managed care organization. Regression analysis was used to evaluate factors influencing hospital charges and length of stay and to adjust for differences in clinical variables between the two provider groups. Hospital charges were primarily determined by length of stay(R2=0.9479, p<0.0001). The academic group cared for 340 babies, while 137 were treated by the private group. The patients admitted by the academic group were more likely to be older (p=0.006), outborn (p<0.0001), require level III care (p<0.0001), suffer from sepsis (p=0.018), require mechanical ventilation (p<0.0001), and have a patent ductus arteriosus(p=0.008), while those admitted by the private group were likely to suffer from transient tachypnea (p=0.003). However, there were no significant differences in outcome parameters between the two provider groups including mortality, age at death, air leak syndrome, bronchopulmonary dysplasia, grade 3 to 4 intraventricular hemorrhage and length of stay. When the data were corrected for differences in risk and patient acuity, however, a significant relationship between physician group and length of stay was observed(p=0.003), with the average length of stay being 11.2 days shorter for the academic group. A net reduction of $4,472,776 in hospital charges might have been realized had the academic group provided care for the 137 private patients. We conclude that: 1) hospital charges in a large children's hospital nursery system are primarily determined by length of stay, 2) in this setting, academic neonatologists produced shorter length of stay and comparable clinical outcome, despite caring for a population at greater risk, 3) the reduction in length of stay may have resulted in a substantial savings in medical care costs. Medical school faculty can provide cost effective health care. Population characteristics must be considered in cost analyses comparing clinical groups and/or strategies.