The alveolate protozoan parasite Perkinsus olseni infects a range of marine bivalves inhabiting tidal flats and shallow subtidal zones, causing considerable damage to shellfish industries. Infection by P. olseni is typically assessed using Ray’s fluid thioglycollate medium (RFTM) assay, with gill tissue often employed as a diagnostic proxy for whole-body infection. However, the diagnostic reliability of gill-based assays across diverse ecological settings—particularly under low-infection conditions—remains uncertain. In this study, we investigated tissue-specific distribution and the diagnostic performance of the RFTM assay in detecting P. olseni in Manila clams (Ruditapes philippinarum) collected from 6 tidal flats along Korea’s west and south coast. The assay was applied to 6 different tissues, including gills and visceral mass. Infection prevalence reached 100% at most sites, except at Padori (90%). Whole-body infection intensity ranged from 0.1×10⁴ to 3.7×10⁶ cells per gram of tissue. The visceral mass consistently harbored the largest proportion of parasites (27.8%–49.0%), followed by the mantle (17.4%–30.6%) and gills (19.4%–25.2%). Gill infection levels correlated strongly with whole-body infection intensity (r²=0.6–0.95), supporting their diagnostic value in high-infection areas. However, at Padori—where infection levels were lowest—the efficacy of the gill assay dropped to 56%, resulting in a 44% false negative rate. These results underscore the limitations of relying solely on gill tissue in low-infection environments and highlight the need for a context-dependent diagnostic approach. A dual-tissue strategy incorporating both gill and whole-body samples is recommended to improve diagnostic accuracy in P. olseni surveillance of Manila clam populations.