Genetic algorithms (GA) have been widely used in quantitative structure-activity/property relationship (QSAR/QSPR) modeling in recent years and have been shown to generate accurate and robust predictions. In a GA, a population of chromosomes is evolved through the processes of random mutation and crossover and evaluated using a fitness function. Here, we will review the basic principles underlying GA and provide a survey of recent applications in QSAR/QSPR, bioinformatics, and in silico drug design, with particular emphasis on the use of GAs in feature selection and dimensionality reduction, model optimization, conformational search, docking, and diversity analysis. © 2014 Springer International Publishing Switzerland. All rights reserved.