Ischemia-reperfusion (I/R) leads to organ injury and organ dysfunction in a variety of clinical disorders. Preclinical investigations examining leukocyte adhesion molecules in I/R provided overwhelming evidence that blocking the function of leukocyte adhesion molecules would be highly effective in improving outcome in clinically relevant diseases. Unfortunately, all 9 of the recently completed phase 2 and 3 clinical trials examining antiadhesion therapy have failed. In this report, we show that a modest increase in ischemic time results in conversion from a CD18-dependent to a CD18-independent injury. This fundamental change in the mechanism of injury can be reduced by inhibition of caspases leading to blockade of apoptosis. Muscle injury resulting from aortic clamping was measured by release of creatine kinase. I/R injury following ischemia of 60 minutes or less and 3 hours of reperfusion was significantly reduced by pretreatment with anti-CD18 monoclonal antibody. However, 90 minutes of ischemia resulted in a marked increase in injury that was not reduced by CD18 blockade. Importantly, the injury resulting from 90 or 120 minutes of ischemia was reduced by the pancaspase inhibitor z-VAD. We propose that the length of ischemia can result in a fundamental change in the mechanism of injury and that all preclinical investigations of I/R must be evaluated with increasing ischemia if they are to model the clinical disease. The result showing CD18-independent I/R injury is not unique; likewise, protection by caspase inhibitors is not unique. However, we show for the first time that caspase inhibitors are effective when CD18 blockade is not.