Photocatalytic water splitting is a potential solution for energy and environmental problems. By using density functional theory method, here, we present an experimentally unexplored two-dimensional (2D) material, palladium thiophosphate (Pd3(PS4)2), as a promising photocatalyst for water splitting. 2D Pd3(PS4)2 is predicted to be exfoliated mechanically from the bulk phase and is dynamically stable. The calculated band gaps for mono-, bi-, and trilayered Pd3(PS4)2 are 2.81 eV (direct), 2.79 eV (indirect), and 2.70 eV (indirect), respectively. Moreover, their band edges straddle between the redox potentials of water. Thus, 2D Pd3(PS4)2 would be a promising photocatalyst for water splitting. Additionally, the strain has significant impact on light absorption and can further improve the photocatalytic capability. Our work expands the family of 2D materials and highlights a new interesting material for solar hydrogen production.