Transcranial magnetic stimulation (TMS) is a patient-friendly stimulation technique of the brain with interesting perspectives. In clinical psychiatry, limited data are available on activity in psychosis and anxiety, but much research has been done in depression. Major concerns on published papers are the inconsistency of used parameter settings, the restraint numbers of patients in randomised trials, the lack of real sham controlled studies and the quasi inexistent reproducibility of results. The most stringent meta-analysis of TMS in affective disorders found a modest, statistically significant antidepressant effect after 2 weeks of daily treatment of high frequency repetitive left dorsolateral prefrontal cortex stimulation. Although most results are rather weak and not convincing enough to promote TMS as evidence-based antidepressive therapy, they show a measurable action that should not be ignored. Preclinical and clinical effects were observed analysing heterogeneous data, and results comparing TMS to electroconvulsive therapy (ECT) in affective disorders are encouraging. Efforts should continue with emphasis on increasing homogeneity and reproducibility in data. Further refinement of stimulation parameters should be established, so that new and large double-blind, long-term, sham-controlled trials can bring us to better understanding and standardising TMS procedure, finally leading to definitive conclusions about its efficacy in psychiatry.