This study was conducted using DSSAT (SUBSTOR-Potato), APSIM and iCrop models for simulating potato growth and yield scientifically comparative assessments of practical strengths and weaknesses. Methodological the approach involves calibration-validation across multiple seasons and fields, RMSE, NSE, and d statistical methods for evaluating crop models applied. The results show that the iCrop model has the highest statistically accurate. It shows that DSSAT and APSIM follow. The iCrop–DSSAT–APSIM ensemble approach is based on the dry yield of the crop and provide the best results in predicting the nitrogen uptake of tuber.