Abstract: In this study for evaluating the irrigation return water, the ground water model was applied to the system. The model calibrated base on prepared conceptual model. Hydrogeological models are suffered from uncertainty and validation. Data scarcity and model calibration difficulties affect the model validation and reality. Due to differences in measurement and element scale of parameter values and variable, the errors associated with input data and boundary conditions and errors of measurement of observation variables, the model validation can't achieve by calibration. For defining all of the parameters, groundwater models need data. Reliability of ground-water model predictions typically depends on several parameters include the correctness of the conceptual model. This study shows the interaction between conceptual model and calibration. This study, relies on the actual situation, shows that the conceptual model could improve during model calibration step.