ABSTRACT. In this work we investigate how the performance of SAT solvers can be improved by SAT formulae configuration. We introduce a fully automated approach for this configuration task, that considers a number of criteria for optimising the order in which clauses and, within clauses, literals, are listed in a formula expressed using the CNF. Our experimental analysis, involving three state-of-the-art SAT solvers and six different benchmark sets, shows that this configuration can have a significant impact on solvers’ performance.