In this study, using the Green-Kubo-method-based molecular dynamics simulations, correlations for predicting the thermophysical properties of nanofluids are developed based on particle shape, fluid temperature, and volume concentration. Silver nanofluids with various nanoparticle shapes including spheres, cubes, cylinders, and rectangular prisms are investigated. The numerical study is conducted within the concentration range 0.14-1.4 vol % and temperature range 280-335 K. The relative thermal conductivity and relative viscosity predicated by the proposed correlations are within a mean deviation of 2% and 5%, respectively, as compared with the experimental results from this study and the available literature. The proposed correlation will be a useful tool for engineers in designing the nanofluids for different applications in industry.