We compared effectiveness (E), the proportion of severely injured patients who were salvageable and survived, to the preventable death rate (PDR) over three consecutive 1-year periods. Severely injured patients were those with at least one injury with an Abbreviated Injury Score (AIS) of greater than or equal to 4. Those with one fatal injury (AIS greater than or equal to 6), a critical head injury (AIS greater than or equal to 5) apart from acute epidural hematoma, or massive multiple injuries (Injury Severity Score greater than 59) were considered nonsalvageable; the remainder were considered salvageable. In the first year, six of 74 salvageable patients died, in the second year five of 76, and in the third year one of 69. The PDR rates were 0.32 (6/19), 0.23 (5/22), and 0.06 (1/17), respectively. There was no significant difference in the E of our trauma program over the 3 years. The apparent improvement in PDR in the second and third years resulted from an increased number of deaths among nonsalvageable patients and fewer deaths among salvageable patients. This finding demonstrates that PDR is sensitive to case mix and not just quality of care, and confirms the superiority of E over PDR for assessing a trauma program.