Mixed states in bipolar disorder have been neglected, and the data concerning treatment of these conditions have been relatively obscure. To address this, we systematically reviewed published pharmacological treatment data for “mixed states/episodes” in mood disorders, including “with mixed features” in DSM–5. We searched PubMed, MEDLINE, The Cochrane Library, clinicaltrials.gov, and controlled-trials.com (with different combinations of the following keywords: “mixed states/features,” “bipolar,” “depressive symptoms/bipolar depression,” “manic symptoms,” “treatment,” “DSM–5”) through to October 2016. We applied a quality-of-evidence approach: first-degree evidence=randomized placebo-controlled studies of pharmacological interventions used as monotherapy; second-degree evidence=a similar design in the absence of a placebo or of a combination therapy as a comparative group; third-degree evidence=case reports, case series, and reviews of published studies. We found very few primary double-blind, placebo-controlled studies on the treatment of mixed states: the preponderance of available data derives from subgroup analysis performed on studies that originally involved manic patients. Future research should study the effects of treatments in mixed states defined using current criteria.