This review by Robert A. McArthur published in Neuroscience & Biobehavioral Reviews presents an overview of some of the pre-clinical and clinical issues that have contributed to the failures of potential novel neuropsychiatric drugs, which have prompted a re-examination of the role of animal models of neuropsychiatric disorders. Publication bias and over-reporting of positive data are discussed among these issues. The author focuses on three factors – technical expertise, statistical expertise and standardization of methodology as the elements needed to improve the reproducibility of preclinical results. In sum, this review nicely illustrates that, while the importance of investing into data robustness and reproducibility is recognized by many experts, this is not the only area neuroscience needs to invest into in order to increase the translatability of preclinical findings.