The successful application of pharmacogenetics in routine clinical practice is still a long way from becoming a reality. In order to favor the transfer of pharmacogenetic results to clinical practice, especially in psychiatry, these studies must be optimized. This article reviews the strengths and weaknesses that characterize pharmacogenetic studies in psychiatry and condition their implementation in clinical practice. We also include recommendations for improving the design of pharmacogenetic studies, which may convert their limitations into strengths and facilitate the implementation of their results into clinical practice. Finally, we discuss the potential value of naturalistic, prospective, multicenter and coordinated projects such as the ‘Phenotype–genotype and environmental interaction. Application of a predictive model in first psychotic episodes’ (known as the PEPs study, from the Spanish abbreviation) in pharmacogenetic studies.