The heat transfer characteristics of R134a during downward condensation are investigated experimentally and numerically. While the convective heat transfer coefficient, two-phase multiplier and frictional pressure drop are considered to be the significant variables as output for the analysis, inputs of the computational numerical techniques include the important two-phase flow parameters such as equivalent Reynolds number, Prandtl number, Bond number, Froude number, Lockhart and Martinelli number. Genetic algorithm technique (GA), unconstrained nonlinear minimization algorithm-Nelder-Mead method (NM) and non-linear least squares error method (NLS) are applied for the optimization of these significant variables in this study. Regression analysis gave convincing correlations on the prediction of condensation heat transfer characteristics using +/- 30% deviation band for practical applications. The most suitable coefficients of the proposed correlations are depicted to be compatible with the large number of experimental data by means of the computational numerical methods. Validation process of the proposed correlations is accomplished by means of the comparison between the various correlations reported in the literature.