PurposeThis paper aims to evaluate the prevalence of self-medication and its associated factors among the Chinese elderly. Also, according to whether the elderly communicate with doctors (no matter before or after self-medication), we aimed to categorize self-medication and explore the associated factors.MethodsIt was a cross-sectional study. Data were derived from the 2018 wave of the China Health and Retirement Longitudinal Study (CHARLS). According to whether communicate with doctors or not, self-medication was reclassified as “self-medicate and NOT communicating with a doctor,” and “self-medicate and communicate with a doctor.” A binary logistic regression was used to identify which elderly were more likely to self-medicate, and a multinomial logistic regression was applied to explore the associated influencing factors of self-medication classifications.ResultsA total of 17,445 individuals aged ≥45 years were enrolled. The prevalence of self-medication was 58.60%. Self-medication was strongly associated with sex, education level, pension, self-reported general health status, chronic illness, satisfaction with local medical services, and three province-level socioeconomic welfare variables. About 19.64% of self-medication populations had communicated with a doctor. Higher education level and younger age were significantly associated with a higher probability of “self-medication and communication with a doctor.”ConclusionThe prevalence of self-medication among the Chinese elderly is increasing over the year. Health education on appropriate medication use targeting elder adults with low education levels is highly recommended. The typology of self-medication and its factors are new research entry points and could be meaningful for future studies.