Intraperitoneal sepsis is difficult to diagnose in thermally injured patients. We reviewed the use of diagnostic peritoneal lavage (DPL) in burn patients suspected of having intraperitoneal infection. Seventeen patients were identified in whom celiotomy, autopsy, or complete recovery could be used to validate the lavage results. A lavage was considered positive if there were greater than 500 white blood cells per mm3 or if microorganisms were present on Gram stain. Six patients had a positive DPL and 11 patients had a negative DPL. There were six true positive, no false positive, ten true negative, and one false negative studies resulting in a sensitivity of 0.86, specificity of 1.00, and diagnostic accuracy of 94%. No complications related to the DPL occurred. This procedure is safe and will rapidly and reliably discriminate between patients needing urgent celiotomy and those requiring further investigation to identify a source of sepsis.