To examine the efficacy and safety of TAS5315, an irreversible covalent Bruton's tyrosine kinase inhibitor, in Japanese patients with rheumatoid arthritis (RA) refractory to methotrexate. In part A of this phase IIa double-blind study, patients were randomised to TAS5315 4 or 2 mg or placebo once daily for 12 weeks; in part B, all patients received TAS5315 for another 24 weeks. The proportion of patients meeting American College of Rheumatology criteria for 20% improvement (ACR20) at week 12 was assessed (primary endpoint). Ninety-one patients were randomised in part A, and 84 entered part B. At week 12, 78.9% of patients achieved ACR20 in the TAS5315 combined group vs 60.0% with placebo (p=0.053), 33.3% vs 13.3% achieved ACR50 (p=0.072) and 7.0% vs 0.0% achieved ACR70 (p=0.294), respectively. More patients receiving TAS5315 than placebo had low disease activity or remission at week 12. Clinical and biomarker improvements were maintained during part B. Adverse event (AE) incidence in TAS5315 was similar to placebo in part A; common AEs with TAS5315 were nasopharyngitis (10.3%), pruritus (6.9%) and cystitis (5.2%). Over 36 weeks, nine patients experienced bleeding events of whom four and two patients recovered with drug continuation and interruption, respectively. Three patients recovered after TAS5315 discontinuation. The primary endpoint was not achieved. TAS5315 appears to have some bleeding risks, but nevertheless demonstrated numerical differences, compared with placebo, in the improvement rates of all measures of RA disease activity. Future analysis of the risk-benefit of TAS5315 should be considered. NCT03605251, JapicCTI-184020, jRCT2080223962.