Objective: To evaluate the efficacy of tocilizumab, an antibody against IL-6 receptor, in patients with hand osteoarthritis. Methods: This was a multicentre, 12-week, randomised, double-blind, placebo-controlled study from November 2015 to October 2018. Patients with symptomatic hand osteoarthritis (pain ≥40 on a 0–100 mm visual analogue scale (VAS) despite analgesics and non-steroidal anti-inflammatory drugs; at least three painful joints, Kellgren-Lawrence grade ≥2) were randomised to receive two infusions 4 weeks apart (weeks 0 and 4) of tocilizumab (8 mg/kg intravenous) or placebo. The primary endpoint was changed in VAS pain at week 6. Secondary outcomes included the number of painful and swollen joints, duration of morning stiffness, patients’ and physicians’ global assessment and function scores. Results: Of 104 patients screened, 91 (45 to tocilizumab and 46 to placebo; 82% women; mean age 64.4 (SD 8.7) years) were randomly assigned and 79 completed the 12-week study visit. The mean change between baseline and week 6 on the VAS for pain (primary outcome) was −7.9 (SD 19.4) and −9.9 (SD 20.1) in the tocilizumab and placebo groups (p=0.7). The groups did not differ for any secondary outcomes at weeks 4, 6, 8 or 12. Overall, adverse events were slightly more frequent in the tocilizumab than placebo group. Conclusion: Tocilizumab was no more effective than placebo for pain relief in patients with hand osteoarthritis.