electric vehicles (BEVs) have been gaining popularity. [9,10] In comparison to BEVs, [11] FCVs get rid of range anxiety and long charging time concerns [12] and have the advantage of subzero environment adaptability. [13] However, their commercial application is still largely hindered by the unsatisfactory fuel cell power density, [14] cost, [15] and durability [16] and undeveloped hydrogen infrastructure.As shown in Figure 1, the proton exchange membrane (PEM) fuel cell stack is the core of FCVs, which consists of several hundred repeated cells connected in series. The power density, that is, the ratio of stack output power to volume or weight is now an important indicator measuring the development level of PEM fuel cells. In general, an increase in power density not only means power improvement along with a more compact cell/stack structure but also cost reduction because of fewer cells needed at a fixed power demand. The New Energy and Industrial Technology Development Organization (NEDO) in Japan set stack power density (including end plates) targets of 6.0 (by 2030) and 9.0 kW L −1 (by 2040). [17,18] Meanwhile, the European Union Fuel Cells and Hydrogen 2 Joint Undertaking (EU-FCH2JU) set a similar goal of 9.3 kW L −1 . [19] At present, the state-of-theart commercial PEM fuel cell stack power density is about Next-generation ultrahigh power density proton exchange membrane fuel cells rely not only on high-performance membrane electrode assembly (MEA) but also on an optimal cell structure. To this end, this work comprehensively investigates the cell performance under various structures, and it is revealed that there is unexploited performance improvement in structure design because its positive effect enhancing gas supply is often inhibited by worse proton/electron conduction. Utilizing fine channel/rib or the porous flow field is feasible to eliminate the gas diffusion layer (GDL) and hence increase the power density significantly due to the decrease of cell thickness and gas/electron transfer resistances. The cell structure combining fine channel/rib, GDL elimination and double-cell structure is believed to increase the power density from 4.4 to 6.52 kW L −1 with the existing MEA, showing nearly equal importance with the new MEA development in achieving the target of 9.0 kW L −1 .