Objectives: Depressive episodes and symptoms are common in bipolar disorder and are often treated with medication. The aim of this literature review is to critically examine the empirical evidence supporting the pharmacological treatment options in bipolar depression. Methods: Randomized controlled trials (RCTs) with parallel-group design of medications for bipolar depression were identified using PubMed. Reference sections were searched for publications of additional RCTs. The quality of the trials was rated. Results: We identified 35 studies. Most trials are of moderately good quality, underpowered and use active comparators of uncertain merit. Large and replicated clinical trials support the efficacy of quetiapine and the combination of olanzapine and fluoxetine. One very large trial found olanzapine alone to be efficacious. One study found lamotrigine to be superior to placebo on some outcome measures, but several negative lamotrigine trials have remained unpublished. The efficacy of lithium is supported by small trials in mixed samples. There is no convincing evidence that antidepressants are effective in bipolar depression, with the possible exception of tranylcypromine. Conclusion: Quetiapine, and olanzapine alone or in combination with fluoxetine, are well documented treatments for bipolar depression. The evidence in favour of lithium and lamotrigine is weaker. Commonly prescribed antidepressants are probably unhelpful.