Two fundamental scaling effects in a proton exchange membrane (PEM) fuel cell system are being discussed—scaling of flow channel size and catalyst particle size. Various micro/nanofabrication processes [lithography, physical vapor deposition, and focused ion beam (FIB) etch/deposition] were employed to produce a microscale experimental platform. In addition, multiphysics fuel cell models were utilized to verify the experimental results. The modeling result suggests that fuel cell power density increases with decreasing channel size due to the reduced diffusion blockage of ribs and increased convection in microchannels. However, experimental observation revealed an optimum channel size with maximum power density. The discrepancy is explained by cathode flooding that the single-phase model is unable to account for. A novel micro-electrochemical impedance spectroscopy technique, combining atomic force microscope with an electrochemical interface, confirmed a linear dependency of Faradic impedance on the triple-phase boundary length (TPB) in microscale catalyst particles. Interestingly, the particle perimeter dependency of Faradic impedance changes to particle area dependency as the particle size increases. This is explained by the generation of cracks in larger catalyst particles, which, in turn, serve as triple phase boundaries. The model prediction based on finite element method confirms the experimental observations.