[en] A set of FORTRAN subroutines is described for calculating water thermodynamic properties. These were written for use in a transient thermal-hydraulics program, where speed of execution is paramount. The choice of which subroutines to optimise depends on the primary variables in the thermal-hydraulics code. In this particular case the subroutine which has been optimised is the one which calculates pressure and specific enthalpy given the specific volume and the specific internal energy. Another two subroutines are described which complete a self-consistent set. These calculate the specific volume and the temperature given the pressure and the specific enthalpy, and the specific enthalpy and the specific volume given the pressure and the temperature (or the quality). The accuracy is high near the saturation lines, typically less than 1% relative error, and decreases as the fluid becomes more subcooled in the liquid region or more superheated in the steam region. This behaviour is inherent in the method which uses quantities defined on the saturation lines and assumes that certain derivatives are constant for excursions away from these saturation lines. The accuracy and speed of the subroutines are discussed in detail in this report. (author)