The trapeziometacarpal joint (TMJ) is the most commonly involved arthritic joint in the hand and is often injected in the outpatient setting. This study assesses the accuracy of TMJ injections. Six pairs of thawed, fresh-frozen cadaveric upper limbs were placed in the anatomic position. The limbs were randomized to be injected by one of two clinicians (a senior and a junior orthopaedic trainee). The TMJ of these specimens was palpated and injected with 0.5mls aqueous jelly dyed with methylene blue. An independent investigator dissected the specimens and the location of the dye was recorded. A Posterior-Anterior radiograph was then taken to assess the bony anatomy of the joint and graded according to Eaton's classification. Dye was found inside the joint capsule in 10 (83%) of the 12 specimens. Using Fishers Exact test no significant difference was found between the 2 injectors (p=0.46). The 2 joints where the dye was extra-articular had grade III and IV arthritis, whereas all other joints were graded I. This study shows that good accuracy of TMJ injection can be achieved using palpation in the earlier stages of TMJ arthritis, when surface anatomy is accurate enough for an intra-articular injection. This is also when synovitis is more prevalent and injections are more relevant. However the failure rate of injections increases as the disease advances.