Background: The pathophysiological processes of bipolar disorder (BD) may be detectable by the use of cerebrospinal fluid (CSF) biomarkers. Aim: We aimed for the first time to review studies of CSF biomarkers in patients with BD compared to healthy control individuals (HC). We investigated the effect of diagnosis, age, gender, clinical state, medication, technical characteristics of tests, fasting state and, cognitive function if applicable. Method: We did a systematic review according to the PRISMA Statement based on comprehensive database searches for studies on cerebrospinal biomarkers in patients with bipolar disorder versus HC. Risk of bias was systematically assessed. Results: The search strategy identified 410 studies of which thirty-four fulfilled the inclusion criteria. A total of 117 unique biomarkers were investigated, out of which 11 were evaluated in more than one study. Forty biomarkers showed statistically significant differences between BD and HC in single studies. Only the findings of elevated homovanillic acid and 5-hydroxy-indoleacetic acid were replicated across studies. Most studies had a cross sectional design and were influenced by risk of bias mainly due to small sample size, lack of data on mood state at the time of the CSF puncture and not considering potential confounders including age, gender, diagnoses, BMI, life style factors such as smoking, and psychotropic medication. Conclusion: Specific monoamine CSF biomarkers may be related to the pathophysiology of BD. Future studies must aim at increasing the level of evidence by validating the positive findings in prospective studies with stringent methodology.