Lithium salts are among the drugs of choice for the treatment of bipolar disorder. Despite six decades of intensive research and an accumulating number of known cellular targets, lithium's mechanism of action still needs to be unraveled. The evolution of large-scale gene-expression analysis methodologies has provided a promising tool to understand the cellular events underlying the mood-stabilizing effect of the drug. However, despite great improvement achieved in transcriptome studies, findings of genes differentially expressed by lithium treatment exhibit, so far, a low reproducibility rate. This review discusses the different design and data analysis strategies applied in the studies and summarizes the possible reasons for the discrepancies among the reports.