The arterial ketone body ratio is calculated as the ratio of arterial levels of acetoacetate/beta-hydroxybutyrate. It correlates with survival in experimental hemorrhagic shock and outcome after liver surgery and myocardial infarction. Procedures for determination of ketone bodies are often laborious and unreliable. As yet the relationship between results from arterial and venous samples is unclear. We therefore describe the determination of the ketone bodies acetoacetate and 3-hydroxybutyrate by an easy, reliable, rapid, inexpensive enzymatic assay using 3-hydroxybutyrate dehydrogenase (E.C. 1.1.1.30) in a semi-automated setting that does not require deproteinization. Preanalytical parameters, including separation from corpuscular elements within 1 h and storage on ice for less than 1 h, must be strictly observed to avoid rapid decay of acetoacetate by spontaneous decarboxylation. The assay has high sensitivity, specificity (+/-5%), and precision (CV <2.5%) with a measurable range of 5-500 micromol/l for either ketone body, and requires only 23.5 microl of plasma. At temperatures below -17 degrees C plasma may be stored for prolonged periods. Results from prospectively scheduled simultaneous sampling of arterial blood and venous blood from the right atrium in 100 consecutive patients with severe multiple trauma (mean Injury Severity Score 38+/-13) support the view that the lung has no role in ketone body metabolism. We conclude that central venous blood can safely be substituted for arterial blood for determination of the ketone body ratio.