Evaluating total antioxidant capacity (TAC) have been playing a vital role in health guidelines, the analysis of nutritional composition in foods, and the diagnosis and treatment of oxidative stress-induced disease. However, numerous antioxidants with similar physicochemical properties make them very difficult to be discriminated because of the similarly analytical signal output. In this work, a novel wavelength-dependent and three-channel photoelectrochemical (PEC) sensor array based on Bi2WO6/Vo-TiO2 electrospun nanostructures was established for the identification of antioxidants in natural teas. The formation of heterojunction interfaces between Bi2WO6 nanosheet and Vo-TiO2 NFs, and the introduction of the oxygen vacancies in TiO2 NFs promoted the visible light absorption ability and electron transfer rate of Bi2WO6/Vo-TiO2 NFs, causing the differences of photocurrent response to different antioxidants under different wavelengths. Furthermore, excited wavelengths, antioxidant types, and photocurrent were regarded as three-factors of fingerprint to accurately discriminate five antioxidants via linear discriminant analysis (LDA). Finally, the PEC sensor array was able to estimate TAC in five actual tea samples (green, yellow, black, white, and dark teas). This development of LDA-assisted multiplex PEC fingerprinting sensor array showed a good promise for the identification and authentication of agricultural products, and the establishment of the database.