Proton exchange membrane, a crucial component of fuel cells, significantly influences both the state-of-health of fuel cell stacks and the safety due to internal gas crossover. Numerous methods exist for detecting hydrogen crossover current, a key functional parameter of membrane, but they are difficult to be applied directly under onboard condition. This study achieves a breakthrough in on-board, in-situ, and quantitative identification of cellindividual hydrogen crossover in fuel cell stacks. The voltage response characteristics in the shutdown process are found to depend on hydrogen crossover, and subsequently an equivalent circuit model is established to identify hydrogen crossover current. The cell-individual hydrogen crossover currents derived from the on-board method correspond with the benchmark detected by the micro-current excitation method, and their variation with anode back pressure aligns with theoretical analysis. Shutdown curves under different hydrogen crossover currents are derived and can facilitate rapid hydrogen crossover estimation and abnormal membrane diagnosis. The shutdown method exhibits robust in repetitive experiments, and possesses good adaptability to noise signals, sparse sampling, and arbitrary auxiliary load control, thereby offering a novel approach for the online assessment of hydrogen crossover current.