In this article, a novel photovoltaic (PV) bypass diode fault detection algorithm is presented. The algorithm consists of three main steps. First, the threshold voltage of the current–voltage (I–V) curve is obtained using different failure bypass diode scenarios. Second, the theoretical prediction for the faulty regions of bypass diodes is calculated using the analysis of voltage drop in the I–V curve as well as the voltage at maximum power point. Finally, the actual I–V curve under any environmental condition is measured and compared with theoretical predictions. The proposed algorithm has been experimentally evaluated using a PV string that comprises three series-connected PV modules, and subtotal of nine bypass diodes. Various experiments have been conducted under diverse bypass diodes failure conditions. The achieved detection accuracy is always greater than 99.39% and 99.74% under slow and fast solar irradiance transition, respectively.