During a cross-validation practice of urinary analysis of pyrethroid insecticide metabolite (3-phenoxybenzoic acid, 3-PBA) by two laboratories (Lab A and Lab B), difference in 3-PBA concentration was noticed. The analytical methods used in the exercise were enzymatic deconjugation (glucuronidase/sulfatase) followed by solid phase extraction and isotope dilution LC-MS-MS determination (Lab A) and acidic deconjugation followed by liquid-liquid extraction and gas chromatography-mass spectrometry determination (Lab B). A significant difference was found for one of the five samples: lower value was obtained in Lab A. Use of acidic deconjugation in Lab A resulted in comparable value with the analytical result for this sample originally obtained in Lab B. The comparison was extended to 42 urine samples taken from Japanese males in Lab A by using the two different deconjugation methods and LC-MS-MS determination. Significantly lower value was obtained for enzymatic deconjugation in some of the 42 urine samples. The results suggested the presence of other conjugated species of 3-PBA than glucuronide and sulfate in human urine. Although the overall agreement between the values obtained by the deconjugation methods was fair, it appears that urine samples should be pretreated by acidic deconjugation for the analysis in biological monitoring of pyrethroid exposure.