The basilar thumb joint is the joint second most commonly affected by osteoarthritis (OA) in the hand. Evaluation of dorsal subluxation of the thumb during a functional task such as key pinch could help assess OA risk. The objectives of this study were to determine the best imaging angle for measuring thumb dorsal subluxation during key pinch and to compare subluxation to corresponding OA grades on the Eaton–Glickel, Outerbridge, and International Cartilage Repair Society scales. Eleven cadaveric forearm specimens were rigged to simulate key pinch. A mobile c-arm captured anteroposterior (AP) view images of the hand and was rotated in 5 deg increments toward the ulnar aspect of the arm up to 60 deg. Dorsal subluxation was measured on each image and compared to determine which angle captured maximum subluxation. The resulting best imaging angle was used for comparisons between dorsal subluxation of the thumb and OA grades for the basilar thumb joint. The max subluxation was in the AP view for most specimens. There was a significant correlation between subluxation and the Eaton–Glickel grade (p = 0.003, R2 = 0.779), but not with either Outerbridge grades (p = 0.8018) or International Cartilage Repair Society grades (p = 0.7001). Our results indicate that dorsal thumb subluxation during key pinch should be measured in the AP view of the forearm/hand. Dorsal thumb subluxation during key pinch had a significant correlation with the Eaton–Glickel radiographic measure of OA but not with more accurate visual classifications of OA.