Music has been demonstrated to improve cognitive test performance in neuropsychiatric populations. However, the impact of music on cognitive training effects, and the importance of music preference, has yet to be studied. This is an essential oversight because many cognitive training programs play music in the background. We sought to determine if participant-preferred or random music would increase the efficacy of computer-based attention training (AT). Forty-eight patients with schizophrenia were randomly assigned to 2 weeks of either: 1) AT with participants' choice of background music, 2) AT with random background music, 3) AT without music, or 4) a no training or music control-watching videos without AT or music. All groups except the no training/no music control group demonstrated improvement in reaction time and response accuracy after training, with those participating in AT with their choice of music exhibiting greater gains than either of the other two AT groups. These findings suggest that complimenting AT with music, and allowing participants to choose the music, may increase the efficacy of AT.