The study areas are located in the western part of Romania, in the historic region Banat, with a relatively varied relief formed of low fields, high fields, piedmonts and hills. The identification of areas with a high probability of new archaeological sites was done through the GIS modeling using complex data sources. The first step was to establish the essential factors in the location and the spatial distribution of known sites. The factors considered are : altitude, slope, aspect and distance from water courses ; the values of these parameters are slightly different from one type of relief to the other. Based on statistic analysis of the values of considered factors for each archeological site, values intervals with the greatest favorability were identified. The second step had to standardize these factors. Because all analyzed factors vary in space, the most appropriate method to standardize is by using Fuzzy membership functions, different for each factor. With their help, these factors were standardized on a scale from 0 (least favorable) to 255 (extremely favorable). Finally, the combination of all these factors allows the identification of high probability areas for archaeological sites. Because the factors considered don’t have the same importance for the favorability of finding a possible site, before the combination of the previous standard factors, each factor was assigned a relative weight to use in the Analytical Hierarchy Process (AHP) implemented by the IDRISI Andes software. The combination of factors considering each weight resulted in the creation of the final model, which presents the probability of identifying new archaeological sites. The probability degree varies continuously in space, from a very high probability (255) to a very low probability (0) depending on the combination mode in a certain area of the considered factors and their weight. Thus generated models, for each relief type, were applied to the archaeological sites from the third and fourth century C. E. and were verified in the field, noting its utility in the identification of new archeological sites.