Under ordinary circumstances choosing the best treatment for each individual patient is a challenging issue. This choice should be based on three main aspects: the patients' psychopathological status, the clinical experience and the best scientific knowledge available regarding efficacy and safety of each treatment. The approach that integrates these three aspects is known as Evidence Based Medicine (EBM). The aim of this report is to present the main methodological features of what constitutes an evidence base, in order to assist mental health professionals in the compelling world of scientific evidence. We describe randomized controlled trials as a tool for evaluating the efficacy and tolerability of treatments, systematic reviews of randomized trials, and meta-analysis, a statistical technique that allows to calculate a weighted pooled estimate of treatment efficacy and tolerability. Finally we will describe two standardized tools for assessing quality of evidence, the CONSORT (Consolidated Standards for Reporting of Trials) statement and the GRADE (Grading of Recommendations Assessment, Development and Evaluation).