We assessed hepatitis E virus (HEV) seroprevalence in patients with hepatic disorders as well as in human immunodeficiency virus (HIV)-infected patients and emphasised the issue of possible non-specific anti-HEV seroresponse and need for combining diagnostic methods for hepatitis E diagnosis. Over a two-year period, from March 2011 to February 2013, we determined anti-HEV immunoglobulin M (IgM) and IgG by enzyme immunoassays (EIA; Mikrogen, Germany) in 504 hepatitis patients negative for acute viral hepatitis A-C. Furthermore, 88 samples from randomly selected consecutive HIV-infected patients were also analysed. All EIA reactive samples were additionally tested by line immunoblot assays (LIA; Mikrogen, Germany). HEV nested reverse transcription polymerase chain reaction (RT-PCR) was carried out in 14 anti-HEV IgM LIA-positive patients. Anti-HEV IgM or IgG were detected in 16.9 % of patients by EIA and confirmed by LIA in 10.7 % [95 % confidence interval (CI) 8.3-13.7 %] of hepatitis patients. HEV RNA was detected in five patients. The agreement between EIA and LIA assessed by Cohen's kappa was 0.47 (95 % CI 0.55-0.75) for IgM and 0.83 (95 % CI 0.78-0.93) for IgG. Anti-HEV IgM and IgG seroprevalence in HIV-infected patients was 1.1 %, respectively. Our findings show a rather high HEV seroprevalence in patients with elevated liver enzymes in comparison to HIV-infected patients. Discordant findings by different methods stress the need to combine complementary methods and use a two-tier approach with prudent interpretation of reactive serological results for hepatitis E diagnosis.