The present study examined potential effect modifiers between polycyclic aromatic hydrocarbon (PAH) exposure and the development of rheumatoid arthritis (RA) and elucidated the relationship between PAHs and RA in subgroups using data from the National Health and Nutrition Examination Survey (NHANES) (2003-2014). The relatedness between eight PAH metabolites and RA in the whole population and different subgroups was tested using multivariable logistic regression analyses. This study included 6297 participants, including 400 RA patients and 5897 non-RA control participants, with full data. Compared to the lowest quartiles, risk of RA was increased in population with the highest quartiles of 1-hydroxynaphthalene (1-NAP), 2-NAP, 2-hydroxyfluorene (2-FLU), and 3-FLU in a bias factor corrected model. The associations between urinary PAH metabolites and RA were prominent in female, young and middle-aged, obese, smoking and alcohol-consuming populations in the subgroup analysis. Our results demonstrated that PAH exposure was related to RA, and the relationship between urinary PAH metabolites and RA differed between subgroups and depended on specific PAH metabolites.