Whipple's disease (WD) is a chronic infectious disorder of multi-system caused by Tropheryma whippelii (TW). Although the prominent symptoms are diarrhea, malabsorption, and weight loss, the arthritis may precede the gastrointestinal symptoms [1]. Furthermore, some cases of WD shows the clinical manifestations of spondyloarthropathy (SpA), such as oligoarthritis pattern, axial involvement, sacroiliitis, and association with HLA-B27 [2].The presence of periodic acid-SchiV-staining foamy macrophages in the lamina propria of duodenum is diagnostic of WD [1]. Recently the molecular diagnosis using TW-speciWc gene ampliWcation is promising. The sequencing of 23S ribosomal RNA (rRNA) gene was established and it provided a basis for molecular diagnosis in the absence of reliable culture and serology [3].As an eVort to disclose whether TW is associated with the pathogenesis of SpA, we investigated the prevalence of TW DNA in saliva of patients with SpA and in healthy controls.A total of 108 patients who were diagnosed as ankylosing spondylitis were enrolled. As a control group, we recruited 132 healthy volunteers. There were no diVerences seen in age distribution and sex between the cases and the controls.Participants were asked to rinse the mouth with tap water to remove any remaining food. Approximately 5 ml of whole saliva was collected by having subjects spit into 50 ml sterile collection tubes. The salivary samples were stored at ¡70°C until DNA was extracted.All DNA samples had been tested using semi-nested PCR for hypervariable region of domain III of the 23S rDNA with primers HGC-23InsF and TW-23InsR1 and for nested re-ampliWcation TW-23InsF and TW23InsR2 were used [3].We found two cases and two controls to be PCR-positive in the saliva. This leads to prevalence rates of 1.5% (95% CI 2.0-7.6%) in controls and 1.9% (95% CI 4.0-11.3%) in cases. In contrary to previous western reports [4,5], the prevalence value of TW DNA in Korean patients without WD is signiWcantly low. This may explain that there is so far no any reported case of WD in Korea. All specimens which PCR was positive were analyzed by direct sequencing in order to conWrm. They were identical to the searched sequence for TW (http://www.ncbi.nlm. nih.gov/entrez/query.fcgi?db=Nucleotide).The previous study showed that all PCR results for 16S rRNA were negative in patients with unexplained seronegative oligoarthritis or polyarthritis, and TW does not seem to be frequently involved in this clinical setting [6].