Shear wave velocity is a basic engineering tool required to define dynamic properties of soils. In many instances it may be preferable to determine Vs indirectly by common in-situ tests, such as the Standard Penetration Test. Many empirical correlations based on the Standard Penetration Test are broadly classified as regression techniques. However, no rigorous procedure has been published for choosing the models. This paper provides 1) a quantitative comparison of the predictive performance of empirical correlations; 2) a reproducible method for choosing the coefficients of previous empirical methods based on the particle swarm optimization and 3) taking into account the polynomial correlation, a new model proposed. Different empirical correlations are compared with different validation criteria. The best performing empirical correlationsresult in a new modeland the unique coefficient associated determined by particle swarm optimization concluded. The more recent correlation only marginally improves prediction accuracy; thus, efforts should focus on improving data collection.