AbstractThe paper illustrates the way in which the principles of statistical mechanics may be applied to the prediction of the thermophysical properties of fluids for the chemical industry. It is shown how the combination of theory with a modest amount of accurate experimental measurement can be a cost-effective replacement for the impractical proposition of the measurement of all properties for all materials of interest and for the unreliable method of totally empirical prediction.Examples of the successes of such an approach are given for the transport properties of the dilute gas, the dense gas, the liquid phase, and the near-critical state of a fluid, as well as for fluid mixtures. Although the rigor of the theoretical approach decreases as the complexity of the molecules increases from simple atomic species, and as the density increases, it is still possible to develop reliable predictive procedures if an appropriate interpretation of the theoretical model is made.