Due to their low volumetric energy density at ambient conditions, both hydrogen and natural gas are challenging to economically store onboard vehicles as fuels [1,2]. One strategy to densify these gases is to pack the fuel tank with a porous material [3]. Metal-organic frameworks (MOFs) are tunable, nanoporous materials with large internal surface areas and show considerable promise for densifying gases [1,[3][4][5][6]. The US Department of Energy (DOE) has set volumetric deliverable capacity targets [7,8] which, if met, would help enable commercial adoption of hydrogen/natural gas as transportation fuels. Many have attempted to establish theoretical upper bounds on the deliverable capacity via pressure-swing adsorption using simplified models of the gas-substrate and gas-gas interaction [9-12]; none have established a rigorous upper bound on the deliverable capacity.Here, we present a theoretical upper bound on the deliverable capacity of a gas in a rigid material via an isothermal pressure-swing. To provide an extremum, we consider a substrate that provides a spatially uniform potential energy field for the gas. Our bound is unique in that it directly relies on experimentally measured properties of the (real) bulk gas, without making approximations. We conclude that the goals set by the US DOE for room-temperature natural gas and hydrogen storage are theoretically possible, but sufficiently close to the upper bound as to be impractical for any real, rigid porous material. However, limitations to the scope of applicability of our upper bound guide fuel tank design and future material development. Firstly, one could heat the adsorbent to drive off trapped, residual gas in the adsorbent [9]. Secondly, the physics of our upper bound do not pertain to any material that changes its structure in response to adsorbed gas, suggesting that flexible materials could still satisfy the DOE targets [13][14][15].