The availability of sample data, together with detailed environmental factors, has fueled a rapid increase in predictive modeling of species geographic distributions and environmental requirements. We founded that MaxEnt model has provided different descriptions of potential distributions based on different sample size, sample accuracy and environmental background. We used six combinations based on three sample data set and two kinds of environmental variables to estimate the potentially suitable areas of Brown Eared Pheasant (Crossoptilon mantchuricum) in MaxEnt model. The results show that the complex variables provided the higher AUC value and accurate potential distribution than simple variables based on the same size of samples. Complicated environmental factors combined with moderate size and accurate sample, can predict better results. The model results were scabrous based on simple environmental factors. Furthermore, big sample size and simple prediction environmental factors will reduce the prediction accuracy, whereas small samples provided a conservative description of ecological niche. Here, we highlighted that considering the big size and high accuracy of sample and many environmental factors of a species to minimize error when attempting to infer potential distributions from current data in MaxEnt model.