One of the early steps in axial turbine design is the use of one-dimensional (1D) mean line calculations to predict the turbine performance and estimate the principal geometric parameters, such as radius and blade heights, that will be needed in further computational fluid dynamic (CFD) studies. This 1D analysis is based on the estimation of the aerodynamic losses expressed as a function of simple blade parameters and the velocity triangles. In this regard, there exist different loss correlations widely used in literature to estimate these losses but at the same time there is a lack of information regarding differentiation between them. Thereafter, the objective in this work was to judge and compare the behaviors of the Kacker- Okapuu, Craig-Cox and Denton loss correlations, all of them widely-used in turbine performance prediction. Present work shows the implementation of these different loss correlations on an existing 1D mean line numerical tool, LUAX-T. Subsequently, once implemented, the correlations were compared and analyzed by the use of a validation process and performing a parametric study. The results show that similar key parameters such as the flow turning, solidity and aspect ratio rule the different loss mechanisms in each correlation. On the other hand, the parametric study shows that the correlations are in agreement with the theory and give similar trends for performance prediction even though they all predict different values of efficiency for the same turbine stage. Moreover, the validation process show the correlations were found to be accurate enough when comparing against two different sets of experimental data. However, it was also proved that the models are only accurate if used within the range of applicability they were developed for, hence a complete knowledge of the limitations of each correlation should be known prior to using them. Finally, the extension of the one-dimensional mean line numerical tool LUAX-T will serve to perform further studies related to turbine design, as there are very few non-confidential turbomachinery design tools available for teaching or researching. Furthermore, a parametric study tool was also developed as part of the program. This last extension and the loss implementation codes are described in this work.