Potomac horse fever is an acute systemic equine disease caused by Ehrlichia risticii. Currently, serologic methods are widely used to diagnose this disease. However, serologic methods cannot determine whether the horse is presently infected or has been exposed to ehrlichial antigens in the past. The purpose of the present study was to compare the sensitivities of the nested PCR and cell culture with that of the indirect fluorescentantibody (IFA) test for the diagnosis of Potomac horse fever. Blood and fecal specimens serially collected from a pony experimentally infected with E. risticii Maryland, blood specimens serially collected from mice inoculated with E. risticii Ohio 380, and blood and/or fecal specimens collected from 27 horses which had clinical signs compatible with Potomac horse fever were examined. These horses resided in Kentucky, Indiana, Pennsylvania, and Vermont. The IFA test titer became positive after 6 days postinoculation (p.i.) for the pony. A culture of the blood of the pony was positive for E. risticii starting on day 1 and was positive through day 28 p.i. By the nested PCR, E. risticii was detectable in the blood and feces of the pony starting on day 1 and was detectable through day 32 p.i. E. risticii was detected in the blood of subclinically infected mice by the nested PCR. Twenty-two clinical specimens were seropositive for E. risticii by the IFA test, with titers ranging from 1:20 to 1:1,280. E. risticii was cultured from 95% (20 of 21) of seropositive clinical blood specimens. E. risticii was detected in the blood by PCR in 81% (17 of 20) of the culture-positive clinical specimens. The study indicated that the nested PCR is as sensitive as culture for detecting infection with E. risticii.