Most industrial areas, especially oilfield operations and geothermal reservoirs, deal with varying viscosities in multicomponent electrolyte solutions. An accurate estimate of this property as a function of pressure, temperature, and varying salt concentrations is highly desirable. Although a number of empirical correlations have already been developed, they are still limited to single electrolyte solutions and can only operate over specified temperature and pressure ranges. In this study, a highly accurate model based on an adaptive network-based fuzzy inference system was developed, mainly devoted to dynamic viscosity prediction in aqueous multicomponent chloride solutions. Crisp input data were transformed into fuzzy sets employing the subtractive clustering algorithm with an effective radius optimized by a hybrid of genetic algorithm and particle swarm optimization technique. Comparing the model with thousands of experimental data concluded in squared correlation coefficient (R-2) of 0.9986 and an average absolute error of 1.59%. The developed model was also found to outperform a number of empirical correlations that are employed for the viscosity determination of single electrolyte solutions.