Background: We have investigated the efficacy of high-frequency left (HFL) versus low-frequency right (LFR) repetitive transcranial magnetic stimulation (rTMS) in depression, focusing on specific symptoms as possible predictors of outcome for these two different types of stimulation. Method: Seventy-four outpatients with a major depressive episode treated with an adequate antidepressant dosage for at least 4 weeks were included in our study and randomly assigned to two different groups: HFL or LFR rTMS. The Hamilton Rating Scale for Depression (HAM-D) items were pooled into 6 factors to evaluate specific symptoms as possible predictors of response. Results: Twenty-one out of 32 patients (65.6%) and 24 out of 42 patients (57.1%) were responders in the HFL and LFR groups, respectively. No significant difference in response rate was observed. Considering the whole sample, we found an inverse correlation between activity and HAM-D score reduction and a significant positive relation between somatic anxiety and outcome. An inverse correlation between psychic anxiety and HAM-D score reduction emerged considering the HFL group. In the LFR group, there was a significant negative relationship between baseline activity and the outcome. Conclusion: These findings support the hypothesis that LFR rTMS could be as effective as HFL rTMS and more suitable for patients with a higher anxiety degree, particularly in bipolar patients.