In this study, a group contribution model is presented for the estimation of the saturated liquid speed of sound of pure chemical compounds. A data set comprised of 1667 experimental data for 74 chemical compounds was extracted from the NIST ThermoData Engine and used to develop and test the model. The least squares support vector machine-group contribution (LSSVM-GC) model uses the occurrences of a set of 43 chemical substructures (to constitute a compound) in addition to temperature to represent/predict the saturated liquid speed of sound. The proposed model produces a low average absolute relative deviation (AARD) of less than 0.6% taking into consideration all 1667 experimental data values.