Measurement of the oxygen metabolite hydrogen peroxide (H2O2) in biological fluids such as plasma could be of interest because it might indicate participation of toxic oxygen species in tissue injury. Recently several reports claimed to measure H2O2 using spectrophotometric and high pressure liquid chromatographic (HPLC) techniques that utilize oxidation of a substrate to a product by a peroxidase. In such a system it is crucial to perform two control experiments to verify whether the measured substance is H2O2. The specificity of the assay for H2O2 should be checked with catalase, and the degradation of H2O2 or inhibition of the assay system by the sample should be checked by determining the recovery of exogenously added H2O2. We performed both types of controls for HPLC and spectrophotometric determinations of H2O2 in plasma and blood. Our results indicate that contrary to previous reports in the literature the measured substance(s) in plasma or blood is not H2O2. Moreover, quantitative measurements of H2O2 in plasma or blood by HPLC was unreliable due to the irreversible binding of H2O2 to the column surface.