During an ongoing epizootic of mycobacteriosis, wild striped bass Morone saxatilis from Chesapeake Bay were analyzed using 3 methods for detection of either mycobacterial infection or associated granulomatous pathology. The specific detection techniques, which utilized aseptically collected splenic tissue, were histology, quantitative culture and nested PCR. Based on analysis of 118 samples, detection of infection differed significantly between the 3 methods (chi-square, p = 0.0007). Quantitative culture and nested PCR detected similar, higher rates of infection (69 and 75%, respectively) than the histological method (52%). Although primary PCR assays for a 924 to 940 bp segment of the mycobacterial 16S rRNA gene were positive for genomic DNA from mycobacterial cultures, a secondary, nested PCR reaction for an internal 300 bp gene segment was required in order to detect mycobacteria within splenic tissue. A similar rate of mycobacterial infection was present in fish collected from all sites tested. Although all detection methods found that striped bass age 4.0 to 4.9 yr had the highest positive incidence, nested PCR detected a higher frequency of mycobacterial infection in fish ≥ 6.0 yr of age than the other 2 methods. Quantitative bacteriology was a more sensitive detection technique when the fish tissue contained ≤10 3 mycobacteria g -1 .