This paper presents an implementation of directly solving Reynolds-Averaged Entropy Transport equation as a part of the CFD solution to predict entropy generation rates in a two-dimensional turbine blade stator section. The Reynolds Averaged Entropy Transport and the necessary modeling. equations are implemented to a commercial CFD solver as a User Defined Scalar (UDS). The results are compared with those obtained by post-processing the temperature and velocity fields obtained by solving full Navier-Stokes equations using a k-epsilon closure. Results show. that there could be significant differences especially in the spatial distributions of entropy fields. This type of approach could offer significant improvements in entropy and loss prediction in turbomachinery flows. However, more studies especially for accurate modeling of temperature fluctuations and temperature-velocity correlations in the entropy transport equation are needed.