OBJECTIVES: The main objective of the study was to categorise trajectories of disease activity measured by DAS28 score of biologic DMARD-naïve (bDMARD) rheumatoid arthritis (RA) patients. The secondary objective was to analyse whether these groups differ in demographic, disease-, treatment-related and cost variables. METHODS: The retrospective analysis used patient-level healthcare utilization data from the National Health Insurance Fund of Hungary. The study period spanned five years from 01.01.2012 to 12.31.2016. The study population included those RA patients, who started their first bDMARD treatment during the study period, had at least two years of follow-up and had DAS28 measurements in every three month. The identification of different trajectory groups based on DAS28 measurements using longitudinal cluster method. RESULTS: The following five groups were separated by the clustering algorithm: initial high DAS28 improving slowly to moderate DAS28, initial high DAS28 improving slowly to low DAS28, initial high DAS28 improving rapidly to moderate DAS28, initial high DAS28 improving rapidly to low DAS28 and initial high DAS28 improving rapidly to remission. The groups were similar in age, gender and in the average time elapsed between diagnosis and the first bDMARD therapy. A significant difference was found between the groups in persistence of the first bDMARD therapy and in the proportion of patients who reached low disease activity and remission in the first half year. There was a notable difference in the burden of disease, and in the proportion of patients using combination therapies. CONCLUSIONS: Distinct disease trajectory groups exist in RA, and there is association between these groups and various patient- and treatment-related variables.