OBJECTIVES: The objective of this study was to describe the real-world use of rituximab, specifically through examining dispensing frequency among patients with rheumatoid arthritis (RA) in England. METHODS: A longitudinal, retrospective cohort study was conducted using the Hospital Treatment Insights (HTI) database, which links pseudonymised patient data with hospital pharmacy dispensing. HTI includes data from January 2010 to September 2014. Included patients were adults with RA, defined using ICD-10 codes, who received ≥4 dispenses (2 courses) of rituximab, with the first course completed within 4 weeks of the index dispense. In line with the Summary of Product Characteristics for rituximab, patients must have received 1000mg at the index dispense. Any patient with a diagnosis of haematological cancer at or prior to index was excluded. RESULTS: One thousand three hundred and eighty-one patients had sufficient data to determine dispensing frequency. The majority of included patients were female (78.9%), aged 53 and above (60.8%), and were dispensed rituximab by a rheumatologist (where consultant type could be identified). Two courses of rituximab was the most common number of courses received, though some patients received up to 7 courses. The median time between courses was found to change with subsequent courses, decreasing from 37.0 weeks between the first and second courses to 23.9 weeks between the sixth and seventh courses. Subgroup analyses performed in 115 patients receiving rituximab in line with National Institute for Health and Care Excellence (NICE) guidance, i.e. after prior biologic therapy, demonstrated that dispensing frequency in this group did not differ markedly from the whole cohort. CONCLUSIONS: The findings of this analysis show that long-term rituximab use in the National Health Service differs from that assumed by NICE in previous appraisals of biologics in RA, with possible implications for future assessments of cost-effectiveness.