BACKGROUND: Diabetes mellitus and alcohol consumption are the most common causes of ketoacidosis in adults. Recently, beta-hydroxybutyric acid (betaHBA) was reported to be a potential serum biomarker in the diagnosis and monitoring of ketoacidosis. We evaluated the performance of T-KB-H and 3-HB kits for the measurement of ketone bodies [acetoacetate (AcAc)+betaHBA] and betaHBA, respectively. METHODS: Quantitative enzymatic assays were performed using the T-KB-H and 3-HB kits (Nittobo Medical Co., Japan) and the Architect ci16200 Integrated System (Abbott Laboratories, USA). Simultaneously, the ketone body levels in these serum samples were determined by gas chromatography-mas spectrometry (GC-MS). We evaluated precision and linearity of these kits and correlation with GC-MS, and established reference intervals in children and adults. RESULTS: The coefficients of variation for the T-KB-H and 3-HB kits were less than 4.0% at analyte levels of 50, 100, and 400 micromol/L. Linearity was observed for AcAc and betaHBA over a 0-1,000 micromol/L range (R2<0.99). Results from the T-KB-H and 3-HB kits were in good agreement with those from the GC-MS analysis, with correlation coefficients of 0.94 for AcAc and 0.96 for betaHBA. Reference intervals determined for the T-KB-H kit were 9.8-270.1 micromol/L and 18.5-531.8 micromol/L in children and adults, respectively. For the 3-HB kit, the reference intervals were 6.4-234.0 micromol/L and 16.0-437.2 micromol/L in children and adults, respectively. CONCLUSIONS: The T-KB-H and 3-HB kits displayed good precision, clinically acceptable linearity, and reliable correlation with an established assay. This indicates that the kits can be used clinically for measuring serum ketone bodies.