The aim of this study was to assess OA prevalence in a cohort designed to study Osteoporosis (OP) using DXA images. Both OA and OP increase with age, however radiographic OA is typically assessed using radiographs and Kellgren–Lawrence grades (KLG), which range from 0 (normal) to 4 (severe). OP is typically assessed using DXA to calculate BMD. Participants were from the North of Scotland Osteoporosis Study (NOSOS), a prospective longitudinal study comprising 1847 women aged 60–82 years, from Aberdeen (N = 1108) and Dingwall (N = 811). DXA images of both hip joints were obtained at baseline using Lunar Prodigy™ scanners (GE Lunar, USA). DXA images for 744 women (1455 hips) from the Dingwall cohort were KLG scored. The score from the hip with the highest KLG score was used to classify OA severity. The threshold for osteoarthritis was set at KL grade ≥ 2 and comparisons between OA, OA severity and baseline variables collected as part of NOSOS were tested using SPSS. A total of 679 women had radiographic OA (91%), with the majority scored at KLG = 2 (76%). Chi-Squared tests showed no significant relationship between self-reported OA and radiographic OA (P = 0.398), or rheumatoid arthritis (P = 0.105). T-tests and Mann–Whitney tests showed that the OA group were significantly older (P < 0.001), shorter (P = 0.005), with higher BMI (P = 0.002) and lower hip and spine BMD (P = 0.007 and 0.004 respectively). This study has shown that DXA images from cohort studies designed to investigate OP can be used to assess the prevalence of radiographic OA to give new insights into the pathology of osteoarthritis.