The development of quantitative structure-activity relationship (QSAR) models for computer-assisted drug design is a well-known technique in the pharmaceutical industry. QSAR models provide medicinal chemists with mechanisms for predicting the biological activity of compounds using their chemical structure or properties. This information can significantly reduce the time to discover a new drug. This work compares and contrasts particle swarms to simulated annealing and artificial ant systems techniques for the development of QSAR models based on artificial neural networks and k-nearest neighbor and kernel regression. Particle Swarm techniques are shown to compared favorably to the other techniques using three classical data sets from the QSAR literature.