Since their appearance in 1993, first approaching the Shannon limit, turbo codes gave a new direction in the channel encoding field, especially since they have been adopted for multiple telecommunication norms. To obtain good performance, it is necessary to design a robust turbo code interleaver. This paper proposes a a differential evolution approach to find above average turbo code interleavers. Performance is compared with the conventional genetic algorithm approach and the empirical results illustrate that DE performs well.