The superiority of graphitic carbon nitride (g-C3N4) nanosheet results from its large specific surface area, which can be achieved by exfoliation of the bulk layered structure. Liquid-phase exfoliation (LPE) is the best-known method for the synthesis of two-dimensional (2D) g-C3N4 nanosheets. However, experimental investigations do not allow for a molecular-level understanding of the process. Molecular dynamics (MD) simulations are expected to provide microscopic insights and quantitative evaluation of the energy consumption during LPE, thus facilitating the search of effective solvents for the LPE of 2D materials.MD simulations are carried out to simulate the LPE process by performing potential of mean force calculations for the separation of two stacked g-C3N4 nanosheets. Free energy of exfoliation is evaluated and compared among nine common solvents with distinct molecular structures.The most probable path for the exfoliation process is identified. The free energy of exfoliation is found to correlate directly with the solvent free energy of a single g-C3N4 nanosheet. The solvation is enthalpy-driven and affected by the mobility of the solvent molecules around the nanosheet. Based on the MD results, several strategies are proposed to guide the selection of solvents for effective LPE.