OBJECTIVES: To evaluate a radiographic atlas for grading foot osteoarthritis (OA) in relation to the relative sensitivity of different radiographic and views and features, and to examine the relationship between radiographic OA and foot symptoms. METHODS: Weightbearing dorso-plantar (DP) and lateral foot radiographs were obtained from 197 people (126 women and 71 men) aged 62-94 years (mean age 75.9, standard deviation [SD] 6.6). The prevalence of OA in five joints (the first metatarsophalangeal joint [1st MPJ], the first cuneo-metatarsal joint [1st CMJ], the second cuneo-metatarsal joint [2nd CMJ], the navicular-first cuneiform joint [N1st CJ] and the talo-navicular joint [TNJ]) was then determined using both views in combination (as recommended in the atlas), or by using either view in isolation. Associations between radiographic OA in individual foot joints and symptoms were then explored. RESULTS: Joint-specific prevalence of OA using both DP and lateral views was 1st MPJ (42.4%), 1st CMJ (22.6%), 2nd CMJ (60.2%), N1st CJ (39.1%) and TNJ (32.7%). Using only the DP view detected almost all cases of 1st MPJ OA (94.6%), however, the sensitivity was lower for the other joints (31.0-60.7%). Using only the lateral view detected almost all cases of OA (83.8 to 86.9%), with the exception of the 1st MPJ and 1st CMJ (50.9% and 60.7%, respectively). Using either osteophytes (OP) alone or joint space narrowing (JSN) alone showed low sensitivity for all joints (14.3-63.0%), with the exception of OP alone in the DP view for the 1st MPJ and JSN in the lateral view for the 2nd CMJ (83.8% and 84.0%, respectively). Radiographic OA in individual foot joints and the total number of joints affected were both moderately associated with foot symptoms. CONCLUSION: Epidemiological and clinical studies should incorporate observation of both OP and JSN from both DP and lateral views to determine the presence of OA in the foot, as the number of cases detected is reduced if only one radiographic feature or view is used. Radiographic foot OA is common in older people and is moderately associated with foot symptoms.