The American College of Surgeons has stated that in considering the development of trauma systems it is important to ensure an appropriate volume of seriously injured patients be seen by each trauma center in order to achieve acceptable mortality rates. Clinical data supporting this recommendation are lacking. An analysis was performed on 1,643 seriously injured trauma patients to determine the relationship between volume and mortality rates. Three separate statistical methods were used: Pearson correlation coefficients, mortality odds ratios, and direct pairwise mortality comparisons. In addition, Tobit analysis was introduced as a method to analyze the relationship between volume and mortality. Mortality rates were adjusted for the confounding variable of serious head injury. Pearson correlation coefficients for volume vs. adjusted mortality was -0.65. Mortality odds ratios comparing the low-volume (less than 140 pts) trauma centers vs. the high-volume (greater than 200 pts) trauma centers was 1.3 for adjusted mortality rates (95% CI = 1.01-1.66; p = 0.04). Categorical analysis showed significantly different mortality rates in the centers before and after adjusting for patient mix. Tobit analysis showed the relationship between volume and mortality to be significant, accounting for 30-40% of the observed variation in mortality rates. In addition, Tobit analysis allowed construction of a model to predict mortality rates, given specific volumes of patients. Our data suggest that an inverse relationship exists between volume and mortality, and support the necessity of configuring trauma systems in a manner that will ensure designated trauma centers will see a high volume of seriously injured patients.