The industrial formulations for the thermodynamic properties of water/steam, which are approximations of the scientific one, are intended to be used in applications where computational speed is of importance, such as in power plant modelling and control. The traditional methods for implementing these tables in software imply either the use of polynomial algorithms, which demand long iteration times, or look-up tables, which require a large memory capacity. On the other hand, there is a group of useful tools, called Artificial Neural Networks (ANNs), that have been successfully applied for pattern recognition and function approximation tasks in, for instance, the areas of medicine, engineering, and economics. This paper aims to show the potential of ANNs for generating the water/steam tables. ANNs enable the production of user-friendly software, which furthermore increases the computational speed while sustaining good accuracy. This new approach avoids the limitations of the traditional methods and can be advantageously implemented in heat and mass balance programs to speed up calculations. Promising results obtained with this technique are highlighted in the present paper, demonstrating the reliability of using ANNs in lieu of polynomials algorithms and look-up tables.