This study aimed to develop prediction models for ionic conductivity and viscosity of ionic liquids (ILs) using quantitative structure property relationships (QSPR) coupled with the descriptors of group contribution type. Utilizing this method, the reverse design of ionic liquids should be available by just a very simple iterative calculation. The polynomial expansion model based on the type of cation, length of side chain, and type of anion was applied to the expression of ionic liquid properties. Parameters of these polynomial expansion models were determined by means of a genetic algorithm (GA), one of the most popular chemoinformatics techniques. We also undertook the reverse design of ILs. Using this group contribution base properties estimation method, our reverse design system cyclopaedically generated lLs structures corresponding to particular physical properties. (c) 2007 Elsevier B.V. All rights reserved.