A considerable amount of literature has recently been published on the study of ionic liquids (ILs). Due to their diversity and to the possibility of being designed to fit a specific application, predictive tools are required to understand their properties and their behaviour when mixed with other substances. One of these tools is thermodynamic modelling with equations of state (EoS). In this work, the Soave-Redlich-Kwong equation of state, coupled to the Debye-Htickel electrolyte model, was applied to correlate the densities of pure ILs. It was assumed that the ILs were totally dissociated, forming a binary mixture of cations and anions. In the modelling, the required EoS parameters were assessed through group contribution methods. One scale parameter was introduced to buffer the uncertainty of the calculations, and it was tuned to fit the model to the ILs' measured densities. Three strategies were adopted to describe the densities in a wide range of temperatures and pressures: (i) adjusting this single IL parameter in order to fit the model to the reported densities at low pressure; (ii) determining it from one measured datum and (iii) from the obtained parameters in the second approach, generating an empirical function that correlates them with their molar masses. At 0.1 MPa (458 data points), the overall absolute average deviation's from the experimental data were 0.9%, 1.3% and 4.0% for each of these strategies, respectively. High-pressure densities were also predicted (3270 points), with computed deviations from the measurements of 1.4%, 1.4% and 3.4%. The proposed method can be used for IL density predictions, as a function of temperature and pressure, when no or only one measured datum is available. (C) 2016 Elsevier Ltd.