An object-oriented and map-based prototype expert system has been developed for integrating geophysical, geological data for base metal exploration. The object-oriented knowledge representation structure and uncertainty propagation mechanisms used work well for this integrated resource exploration problem. Evidential belief function theory is utilized to manage the uncertainties in the system. It appears a more natural and adequate theoretical basis for representing spatially unbalanced geophysical and geological information. The problem of dependent information can be dealt with in a knowledge-based system of this type by explicitly introducing important uncertainties and by organizing the relation network properly. The prototype system has been tested using real exploration data sets.