Pathogen surveillance is common in chelonians and multiple anatomical sampling sites are used for pathogen detection, but agreement between these sites has not been assessed. The objective of this study was to assess agreement between three sampling sites, oral swab (OS), cloacal swab (CS), and combined oral/cloacal swab (OCS), for detecting three pathogens in eastern box turtles (Terrapene carolina carolina). Box turtles (n=88) were assayed for Terrapene adenovirus 1 (TerADV1), Terrapene herpesvirus 1 (TerHV1), and Mycoplasmopsis sp. using quantitative PCR. Agreement between pathogen status based on sampling site was assessed using the Cohen kappa. Agreement was highest for Mycoplasmopsis sp. between OCS and OS (k=0.941), whereas moderate and minimal agreement were noted between OCS and CS (k=0.64) and OS and CS (k=0.538). For TerADV1, agreement was weak between OCS and OS (k=0.559), minimal between OS and CS (k=0.283), and absent between OCS and CS (k=0.204). TerHV1 agreement was moderate between OCS and OS (k=0.783) and absent between OCS and CS (k=0.106) and OS and CS (k=0.052). All pathogens were most frequently detected in OCS samples and DNA concentrations differed between sampling sites (P<0.0001). If testing multiple samples is not possible, OCS sampling improves the detection of these three pathogens over OS and CS alone.