Abstract The hopes for readily implementable precision medicine are high. For many complex disorders, such as bipolar disorder, these hopes critically hinge on tangible successes in pharmacogenetics of treatment response or susceptibility to adverse events. In this article, we review the current state of pharmacogenomics of bipolar disorder including latest results from candidate genes and genome-wide association studies. The majority of studies focus on response to lithium treatment. Although a host of genes has been studied, hardly any replicated findings have emerged so far. Very small samples sizes and heterogeneous phenotype definition may be considered the major impediments to success in this field. Drawing from current experiences and successes in studies on diagnostic psychiatric phenotypes, we suggest several approaches for our way forward.