Ionic liquids (ILs) have gained considerable attention in view of their potential use in various industrial applications to circumvent the setbacks of volatile organic solvents. Generally, the characteristic properties of ILs vary with different combinations of cations and anions. The structures of ILs directly impact upon their properties, in particular the transport properties, such as viscosity. The desired viscosity of ILs, thus, can be achieved by formulating proper combinations of cations and anions. In the present study, the viscosity of different ILs has been estimated and correlated via the Quantitative Structure Property Relationship (QSPR) approach with a set of descriptors, namely interaction energies of the IL cation-anion pairs generated by the Conductor-like Screening Model for Real Solvents (COSMO-RS). Experimental data of imidazolium-based ionic liquids collected from the literature with viscosities in the range of 5.3-6410 mPa s at temperatures in the range of 273.15-393.15 K were used to develop the model. A multilinear relationship between the viscosity and interaction energies using a stepwise model-building approach was applied to generate the correlation model. The proposed QSPR model produced a low average absolute relative deviation (AARD) of 4.66, root mean square error (RMSE) of 0.26 and coefficient of determination (R2) of 0.965, suggesting that the proposed model fits well with the available training set data and can be further validated with a larger data set. The proposed model suggests that the viscosity is highly dependent on van der Waals forces and temperature, but is also controlled by electrostatic and hydrogen-bonding interactions to a minor extent. The obtained results in this work will pave the way for selecting proper ionic liquids with desired viscosity for particular applications. This journal is © 2019 The Royal Society of Chemistry and the Centre National de la Recherche Scientifique.