ABSTRACTSTo determine the sensitivity and specificity of detecting urinary medium-chain acylearnitines for the diagnosis of MCAD deficiency, 114 urine specimens from 75 children with metabolic diseases and controls were analyzed in a blinded fashion using a radioisotopic ex-change/HPLC method. All 47 patients with MCAD deficiency were correctly diagnosed using the criterion hexanoylcarnitine or octanoylcarnitine peak areas larger than those of other medium-chain acylcarnitines. The majority of them were tested during the asymptomatic state without L-carnitine loading. Four patients with other defects of fatty acid oxidation and three patients receiving vatproic acid had a similar acylcarnitine exeretion pattern. To farther examine the specificity of the method, eight infants receiving a diet enriched with medium-chain triglycerides and 13 additional patients receiving valproic acid were studied. Most of these also tested positive for MCAD deficiency by the above criterion. Analysis by a new gas chromatographic-mass spectrometric procedure revealed that octanoylearnite, not valproylcarnitine, was the most abundant medium-chain carnitine ester excreted by a patient treated with valproic acid. Quantitation of urinary hexanoylcarnitine and octanoylcarnitine showed considerable overlap among patients with MCAD deficiency and those receiving valproic acid or a medium-chain triglyceride-enriched diet. MCAD deficiency can be reliably detected in urine specimens by this method without the need for prior carnitine loading. However, other defects in fatty acid oxidation must be differentiated from MCAD deficiency, and a history of medium-chain triglyceride or valproic acid administration must be considered if the diagnosis of MCAD deficiency is sought through analysis of urinary acylcarnitines. (Pediatr Res31: 545–551, 1992)