In antioxidant activity testing, it has been argued that assays capable of measuring the inhibitive action against the biologically relevant peroxyl radicals (ROO center dot) from a controllable source are preferable in terms of simulating physiological conditions because ROO center dot is the predominant free radical found in lipid oxidation in foods and biological systems. A new fluorescent probe, p-aminobenzoic acid (PABA), was developed for selective measurement of peroxyl radical scavenging (PRS) activity of biological samples, in view of the fact that the existing PRS assays are quite laborious and require the application of strictly optimized conditions. The earlier probe, beta-phycoerythrin, of a similar PRS assay of wide use, oxygen radical absorbance capacity (ORAC), varies from lot to lot of production, undergoes photobleaching, and interacts with polyphenols via non-specific protein binding, while the current probe, fluorescein, undergoes undesired fluorescence (FL) quenching and side reactions. The developed technique is based on the fluorescence decrease of the PABA probe (within an optimal time of 30 min) because of its oxidation by ROO center dot, generated from the thermal dissociation of 2,2'-azobis(2-methylpropionamidine) dihydrochloride (AAPH). In the absence of the scavenger, ROO center dot reacted with the probe, generating non-fluorescent products, and caused a decrease in PABA fluorescence, whereas the ROO center dot scavenger resulted in a fluorescence increase because of the inhibition of the probe oxidation by ROO center dot. Thus, the fluorescence increment of intact PABA is proportional to the ROO center dot scavenging activity of samples. The linear range of relative fluorescence intensity versus the PABA concentration was in the interval of 0.5-5.0 mu M. Assay precision and accuracy were assessed by analyzing two spiked homogenates of liver and kidney at clinically relevant concentrations with 97-105% recovery and 2.3% interday reproducibility. The proposed method was successfully applied to assay the ROO center dot scavenging activity of some amino acids, plasma and thiol-type antioxidants, and albumin, with the latter showing the strongest activity with respect to both ORAC and developed PABA methods. On the other hand, the original ORAC method suffers a limitation from protein thiols in total radical-trapping antioxidant parameter (TRAP) calculations, and inconsistent results have been reported by various researchers for ORAC values of thiols, such as vastly differing values for glutathione and zero value for cysteine.