Transformation of programs among different paradigms has been widely studied in academic research and education. The interest on Datalog has recently increased as a specification language for expressing, in just a few lines, complex interprocedural analyses involving dynamically created objects. In real-world problems, the Datalog rules encoding a particular analysis must be solved generally under a huge set of Datalog facts that are automatically extracted from the analyzed program (e.g. pointer dependencies). In that context, this work aims at exploiting Maude’s capabilities for supporting efficient evaluation of Datalog queries. We demonstrate how, starting from an almost straightforward transformation of Datalog programs into Maude specifications, we are able to achieve a highly efficient version.