Molecular dynamics simulation is a powerful tool for understanding and predicting thermophysical properties in systems both at equilibrium and away from equilibrium and composed of molecules ranging in complexity from spherically symmetric monatomic molecules to polymers. The widespread utilization of low-cost high performance engineering workstations has permitted the application of molecular dynamics and Monte Carlo simulation techniques to proliferate at an astounding rate. High-performance massively-parallel supercomputers (such as the Intel Paragon and IBM SP/2) permit the simulation of systems with an unprecedented degree of complexity. As computing costs continue to decline, molecular simulation of systems modeled to a high degree of accuracy promises to become an equal partner with experiment for the prediction of thermophysical properties for industrial purposes. Several examples of the use of molecular dynamics simulation to predict the properties of highly realistic systems are described.