A new approach to constructing point charge water models is presented, which focuses on optimizing the charge distribution to best describe the electrostatics of the water molecule. This method avoids traditional geometric constraints and instead searches for the optimal charge distribution in a 2D parameter space of key multipole moments. The resulting "Optimal Point Charge" (OPC) water model significantly improves the accuracy of bulk water properties compared to commonly used rigid models, with an average error relative to experiment of 0.76%. The model also shows close agreement with experimental data across a wide range of temperatures. The OPC model outperforms other rigid models in predicting hydration free energies of small molecules, with root-mean-square errors less than 1 kcal/mol. The model is based on a 4-point rigid structure and is designed to accurately reproduce the three lowest order multipole moments of the water molecule. The model's parameters are determined through analytical equations that relate the charge distribution to the multipole moments. The OPC model is included in the solvent library of the Amber v14 molecular dynamics software package and has been tested in GROMACS 4.6.5. The computational cost of running simulations with the OPC model is similar to that of the popular TIP4P model. The model's performance is validated through a range of simulations and comparisons with other commonly used rigid models, demonstrating its improved accuracy in reproducing both bulk properties and hydration free energies. The approach presented here offers a more accurate and efficient method for constructing classical water models, with potential applications in developing more accurate models with different numbers of point charges and flexible or polarizable models.A new approach to constructing point charge water models is presented, which focuses on optimizing the charge distribution to best describe the electrostatics of the water molecule. This method avoids traditional geometric constraints and instead searches for the optimal charge distribution in a 2D parameter space of key multipole moments. The resulting "Optimal Point Charge" (OPC) water model significantly improves the accuracy of bulk water properties compared to commonly used rigid models, with an average error relative to experiment of 0.76%. The model also shows close agreement with experimental data across a wide range of temperatures. The OPC model outperforms other rigid models in predicting hydration free energies of small molecules, with root-mean-square errors less than 1 kcal/mol. The model is based on a 4-point rigid structure and is designed to accurately reproduce the three lowest order multipole moments of the water molecule. The model's parameters are determined through analytical equations that relate the charge distribution to the multipole moments. The OPC model is included in the solvent library of the Amber v14 molecular dynamics software package and has been tested in GROMACS 4.6.5. The computational cost of running simulations with the OPC model is similar to that of the popular TIP4P model. The model's performance is validated through a range of simulations and comparisons with other commonly used rigid models, demonstrating its improved accuracy in reproducing both bulk properties and hydration free energies. The approach presented here offers a more accurate and efficient method for constructing classical water models, with potential applications in developing more accurate models with different numbers of point charges and flexible or polarizable models.