Direct measurement of soil water infiltration can be laborious, time-consuming, and expensive under certain conditions. The main objective of this study was to investigate the possibility of accurately estimating the final soil water infiltration rate (IR) using pedotransfer functions (PTFs) based on readily available soil attributes (RASAs). In this study, new PTFs were derived from a small regional dataset, and the performance of existing PTFs was evaluated in comparison with the locally developed ones. Furthermore, an approach was proposed to improve IR estimation by developing error functions based on basic soil properties and adding the predicted error to the estimated IR values. In addition, a recalibration approach was applied to enhance the performance of existing PTFs at the local scale. To this end, IR, soil texture, bulk density, organic matter, geometric mean diameter, geometric standard deviation of particle size, specific surface area, and structural stability index were measured in two adjacent regions (T1 and T2). The IR data were obtained using the double-ring method at 49 points in T1 and 27 points in T2, with three replications. Using the T1 dataset, PTFs were developed through multiple linear regression (MLR), and both new and existing PTFs were validated using the T2 dataset. . The results showed that none of the considered PTFs could accurately estimate IR (34%<nRMSE<5678%) but the use of error functions (25%<nRMSE<104%) and recalibration of the existing PTFs (18%<nRMSE<30%) considerably increased the accuracy of the IR estimations. Consequently, even with a limited number of data (N=15), one can develop error functions or recalibrate the existing PTFs to improve their estimation accuracy in practice.