Background Recent studies have identified chewing of khat (Catha edulis) as an independent risk factor for liver injury; however, the pathogenetic mechanism remains poorly understood. Case series have found markers of autoimmune hepatitis in patients with khat-related liver disease, suggesting that khat chewing might trigger an autoimmune response. The aims of the present study were (i) to assess the prevalence of autoantibodies typical for autoimmune liver diseases in a healthy population in Ethiopia and (ii) to explore the hypothesis that khat usage triggers autoimmunity. Methods Consenting adults (≥18 years) without known autoimmune disease or manifest liver disease were included. One-hundred-and-sixty-nine individuals with current khat use were compared to 104 individuals who never used khat. Seroprevalence of antinuclear (ANA), antismooth muscle (SMA), and antimitochondrial antibodies (AMA) were determined and compared between the groups using logistic regression models to adjust for age and sex. Results Overall, 2.6% of the study subjects were positive for ANA, 15.4% for SMA, and 25.6% for AMA. When comparing khat users to nonusers, ANA was detected in 4.1% vs. 0% (p=0.047), SMA in 16.0% vs. 14.4% (p=0.730), and AMA in 24.9% vs. 26.9% (p=0.704). ANA was excluded from multivariable analysis since there was no seropositive in the reference group. After adjusting for sex and age, no significant association between khat use and SMA or AMA was found. Conclusions No association between khat usage and the seropresence of SMA or AMA was found, weakening the hypothesis that khat-related liver injury is mediated through autoimmune mechanisms. However, the seroprevalences of AMA and SMA were strikingly high in this Ethiopian population compared to global estimates, suggesting that diagnostic algorithms for autoimmune liver diseases developed in Europe and North America might lead to misdiagnosis of patients on the African continent.