Nowadays, following the technological advances and increased accuracy of remote sensing and satellite imagery capabilities, it seems that the combination of these capabilities with the available information of changes or destruction of the resources in the past and present, and base on the modeling will be able to predict the land cover changes in the natural and anthropogenic areas such as protected area. One of the most important ways to obtain this object is the spatial modeling of land cover changes using various strategies. Therefore, the present study was to evaluate the ability of Ca-Markov model to estimate the land cover changes in protected area located in the Bijar state from 1987 to 2013 period and also modeling of this changes and predict of them for 2026. For this purpose, Landsat images of 1987, 2000, and 2013 were used to extraction of land cover maps. Validation of the model was taken by comparing the predicted map of 2013 using model and land cover map for the same year. The rangelands and agriculture lands use in 2013 were about 21180.65 and 10273.53 hectares which prediction result of the model showed 21935.97 and 9631.62 hectares, respectively. This result indicates the high accuracy of the Ca-Markov model to predict the land cover changes in 2013. The results showed that from 1987 to 2000, about 12 percent, and from 2000 to 2013, about 10 percent of the rangelands converted to agriculture lands. The prediction result of Ca-Markov model shows that if the current trend continues, between 2013 and 2026, about 13 percent of the rangelands will converted to agriculture land.