Soil water retention curve in many studies, including simulating the movement of water and solutes, irrigation and drainage issues is required. This soil hydraulic property was determined by direct (field and laboratory) and indirect methods. Direct methods are costly and time consuming therefore the indirect method such as pedotransfer functions (PTFs) are the alternative to estimate this property. PTFs are developed to estimate drying branch of soil water retention curve whereas little research has been done for wetting branch of soil water retention curve. In this study, Shani field method was used to obtain wetting branch of soil water retention curve in agricultural area in 39 site of Boinzahra county.  At the first, 10 national and international parametric PTFs were evaluated to estimate wetting branch of soil water retention curve. Results showed that simulated soil water retention curve using PTFs were not consistent with the measured soil water retention curve. In order to improve simulation of wetting branch of soil water retention curve, non-linear regression is used between simulated and measured value. The results showed that the using derived non-linear regression causes the Root Mean Square Error (RMSE) of PTFs of Mayr-Jarvis (1999), Wosten et al. (1999), Campbell (1985), and saxton et al. (1986) decreases 174, 140, 132, and 113 percent, respectively and consequently improves estimation of wetting branch of soil water retention curve. The results also showed that the using derived non-linear regression causes Geometric mean Error Ratio (GMER) closes to one and prevents over and under estimation of PTFs.