Several types of descriptors were used to derive quantitative structure–property relationship (QSPR) for prediction of surface tension of 142 alcohols at 298.15 K. Genetic function approximation (GFA) was employed to select optimal subset of descriptors that have significant contribution to the surface tension of the training set. By using GFA, a statistically significant QSPR model with the squared correlation coefficient values of 0.9784 for training set and 0.9802 for testing set was developed. On the same data set, another QSPR model was developed based on adaptive neuro-fuzzy inference system (ANFIS) with the squared correlation coefficient values of 0.9882 for training set and 0.985 for testing set. The obtained results in this paper suggest that with the proposed methods, it is possible to obtain a good estimation of surface tension of alcohol compounds. Copyright © 2011 John Wiley & Sons, Ltd.