AbstractThree methods were compared for determining anti‐HAV of the IgM class. In the first method flat‐bottomed microtiter plates coated consecutively with anti‐HAV of the IgG class and HAAg were incubated with patient serum and, after washing, peroxidase conjugated anti‐m̈ was added. After subsequent incubation with substrate the enzymatic reaction was stopped and the optical density was measured. In the second method the solid phase was coated first with antibodies to IgM and after incubation with patient serum and subsequent incubations with HAAg and125I anti‐HAV of the IgG class radioactivity was counted. These two methods were compared with reorienting sucrose gradient ultra centrifugation, an established method for demonstrating specific IgM antibodies. The persistence of IgM anti‐HAV in 103 sera drawn at different times after onset of jaundice was evaluated. Sera drawn up to 30 days after onset of hepatitis A were IgM anti‐HAV positive with both of the first two methods. Forty‐one to 90 days after onset of illness IgM anti‐HAV could be demonstrated with the first method in 47% of the patients, in 94% with the second method, and in 82% with gradient centrifugation. The second method was most sensitive and could be adjusted so that at a serum dilution of 1:104anti‐HAV IgM was detected only up to six months after infection. In contrast to the first method, nonspecific reactions caused by rheumatoid factor were not detected with the second method. During a one‐year period about 15,000 sera of patients with clinical diagnoses of acute hepatitis were tested; the positive results correlated well with the clinical data, and there was no indication of nonspe