Background and ObjectivesThis systematic review with meta-analysis (PROSPERO: CRD42021249336) was performed to estimate the pooled lifetime prevalence of bipolar symptoms (BS) and bipolar disorder (BD) in people with epilepsy (PWE).MethodsA search was performed on June 5, 2021, in 4 databases (MEDLINE/PubMed, Ovid EMBASE, Ovid APA PsycInfo, and Web of Science) for original research reporting on BS/BD in PWE, with no restriction on language or time of publication. Inclusion criteria were as follows: (1) original research, (2) cross-sectional study design component, (3) reported lifetime prevalence of BS/BD or enough information to calculate an estimate, and (4) reported the method by which participants were deemed bipolar. Studies based on an exclusively pediatric population were excluded. To calculate pooled lifetime prevalence of BS/BD, 2 meta-analytic random-effects models were fitted, one for BS and the other for BD. Risk of bias was assessed using a standardized appraisal tool for studies reporting prevalence. Certainty of evidence was evaluated using the Grading of Recommendations, Assessment, Development and Evaluations (GRADE) approach.ResultsA total of 750 records were screened and 17 studies were included for analysis: 7 provided prevalence estimates for only BS, 8 for only BD, and 2 for both BS and BD. After outlier exclusion and subgroup analysis using screening method as a moderator, the pooled prevalence of BS in PWE was 12.3% (95% CI 10.6%–14.1%) (7,506 PWE). The pooled prevalence of BD in PWE was 4.5% (95% CI 2.2%–7.4%) (48,334 PWE). Considerable heterogeneity was present, more so for BD than for BS, and could be explained through differences in population demographics and study methodology.DiscussionThis study's main limitation was regarding the certainty of evidence. However, our estimates of prevalence should prompt further research on BS/BD in PWE. Given the significant morbidity associated with BD, clinicians should carefully screen PWE for BS.