This report presents statistical tests for assessing the accuracy and validity of two regional models of the relationship between dissolved-solids concentration and specific conductance. These models are used to estimate dissolved-solids concentrations based on specific-conductance measurements. The two regional models are compared with station models for 27 stream sites in the Susquehanna River basin. The average standard errors of estimate of two regional models, when used to estimate dissolved-solids concentrations for the 27 stream sites, are 10.7 and 10.6 percent. In comparison, the station models have average standard errors of 9.0 and 9.3 percent, respectively. Results of the statistical tests indicate that in the Susquehanna River basin the regional models can be used to estimate dissolved-solids concentrations from specific-conductance data for sites for which chemical analyses are not available.