Base deficit has been established as a predictor of mortality and endpoint of resuscitation. We hypothesized that in a significant subset of surgical intensive care patients, base deficit is secondary to hyperchloremic acidosis, and that these patients experience lower mortality than those patients whose base deficits are secondary to other causes. Seventy-five consecutive surgical intensive care patients with base deficits greater than 2.0 were prospectively studied. The etiology of the patients' base deficits was determined by admission laboratory data. Patients were divided into those with hyperchloremic acidosis, and those with acidosis from other causes. Mortality within these groups was compared by Fisher's exact test. Thirty-seven patients (49.3%) had hyperchloremic acidosis. Thirty-three patients (46.7%) had lactic acidosis. Three patients (4%) had base deficits secondary to ketosis, and two patients (2.6%) had base deficits secondary to uremia. There were no significant differences in age, APACHE II scores, or volumes of resuscitation between the hyperchloremic group and the remaining patients. There were four deaths (10.8%) in the hyperchloremic group and thirteen deaths (34.2%) in the remaining patients (P = 0.03). Hyperchloremic acidosis resulted from resuscitation with lactated Ringer's solution in 18 (48.6%) of the hyperchloremic patients. Hyperchloremic acidosis is a common etiology of base deficit in the surgical intensive care unit. It is associated with lower mortality than base deficit secondary to other causes. Moreover, it is frequently induced following resuscitation with lactated Ringer's solution. Failure to properly diagnose this subset of acidotic patients may result in inappropriate clinical interventions due to the erroneous presumption of ongoing tissue hypoxia.