Purpose: To assess the efficacy of adalimumab (Humira, AbbVie Pty Ltd) versus placebo in patients with erosive hand osteoarthritis (OA). Methods: This was a single centre randomised double-blind placebo-controlled crossover trial. Patients greater than 50 years old, meeting the ACR criteria for hand OA, with pain >50 on 100 mm Visual Analogue Scale (VAS), morning stiffness >30 minutes and erosive hand OA present on x-ray were recruited. Patients with systemic inflammatory joint disease or known contraindications to adalimumab treatment were excluded. Eligible patients then underwent a 1.5T non-contrast MRI scan of the most painful, erosive hand joint to assess inflammatory OA defined by the presence of synovitis (plus or minus bone marrow lesions) (BMLs) in the index joint. Patients with synovitis were enrolled in the study. Patients were randomised to adalimumab (40 mg subcutaneous injections every other week) or placebo for 12 weeks (treatment period 1) followed by an 8-week washout and then the converse treatment for 12 weeks (treatment period 2). Hand pain (0–500), function (0–900) and stiffness (0–100) were measured using the Australian/Canadian Hand OA Index (AUSCAN) and hand pain using a VAS at baseline, 4, 8 and 12 weeks of each treatment period. Synovitis (grade 0–3) and BMLs (grade 0–3) were scored at baseline and 12 weeks of each treatment period, blinded to treatment allocation, according to the OMERACT hand OA MRI score (HOAMRIS) or the OMERACT thumb base OA MRI scoring system (TOMS). The primary outcome was change over 12 weeks as assessed by the AUSCAN pain subscale. Secondary outcomes included change in the AUSCAN function and stiffness subscale, and VAS pain from baseline to 4, 8 and 12 weeks, change in synovitis and BMLs from baseline to 12 weeks and change in AUSCAN pain subscale from baseline to 4 and 8 weeks. Analysis was by intention to treat. Treatment effects were calculated using paired t-tests, based on within-individual differences over both treatment periods. This accounted for the crossover nature of the study design, that is, each patient served as his or her own control. Results: 51 patients were recruited and 43 were randomised (Group 1 N = 18, active then placebo; Group 2 N = 25, placebo then active). Mean age was 61 years (SD 8.4), 77% were women, and mean AUSCAN pain level was 318.3 (SD 96.8) out of 500 at baseline indicating highly symptomatic disease. At 12 weeks there was no difference between the groups on the primary outcome measure (Figure): mean decrease in AUSCAN pain subscale score was 19.6 (SD 82.6) following adalimumab treatment versus 4.9 (SD 144.5) following placebo treatment (treatment effect 14.7, P = 0.58). No statistically significant differences were found for any of the secondary outcomes. The rate of adverse events was similar in the two groups. There was little change in synovitis and BML scores in both groups and no significant difference between the groups was observed (18% improved following adalimumab treatment vs 16% following placebo treatment). Conclusions: Adalimumab was no different to placebo for pain, synovitis or BMLs in patients with erosive hand OA.