Sub-Neptunes (radii of 1.25-4 Earth radius) remain the most commonly detected exoplanets to date. However, it remains difficult for observations to directly “see” whether these intermediate-sized exoplanets have surfaces and where their surfaces are located. Here we propose a new technique that uses solely the abundances of atmospheric trace species in the visible atmospheres as proxies for identifying surfaces on sub-Neptunes. As an example, we used a state-of-the-art photochemical model to simulate the atmospheric evolution of K2-18b and investigate its final steady-state composition with surfaces located at different pressure levels. We find that surface location has a significant impact on the atmospheric abundances of trace species, making them deviate significantly from their thermochemical equilibrium and “no-surface” conditions. For an assumed H2-rich atmosphere for K2-18b, we identify seven chemical species that are most sensitive to the existence of surfaces: ammonia (NH3), methane (CH4), hydrogen cyanide (HCN), acetylene (C2H2), ethane (C2H6), carbon monoxide (CO), and carbon dioxide (CO2). We propose a flowchart that uses species abundance ratios to help distinguish the existence of surfaces and the rough surface locations. This framework can be applied together with future observations to other sub-Neptunes of interest.