The general equation of state employed in direct form, and for the first time presented in Part I of this series, is extended to correlate and predict the viscosity of binary mixtures of ionic liquids. The extension, also done for the first time, considers the classical mixing and combining rules commonly used in studies on pressure-temperature-volume equations of state, including two cases: (i) with no-interaction parameter (predictive model); and (ii) with one adjustable interaction parameter (correlating model). Data on viscosity of binary mixtures of ionic liquids have been gathered from the literature, analyzed, and selected, to finally construct a database of consistent data to obtain a general model that relates viscosity with pressure, temperature and concentration. A total of 2520 data points distributed on 344 isotherms, for 32 mixtures were considered for analysis, with average absolute relative deviations below 6%. If no interaction parameter is used, the model is still capable of predicting the viscosity of a mixture using only properties of the pure components. Results show that the model is accurate enough, being simpler and more accurate than sophisticated multiparametric models. (C) 2019 Elsevier B.V. All rights reserved.