OBJECTIVES: To develop an electronic version of the Health Assessment Questionnaire Disability Index (HAQ-DI) for use on a handheld computer and to assess the equivalence of the electronic and paper versions in patients with rheumatoid arthritis. METHODS: Development of the electronic version involved significant modification, in particular eliminating free text entry. The conceptual framework of the scale was not altered, and the electronic version was aligned to the original scoring scheme. Equivalence of scores was evaluated in a crossover study in which patients completed a series of questionnaires including the HAQ-DI once in electronic and once in paper mode in randomised order with a 45-minute interval in between. Patients then completed a questionnaire on ease of use and acceptability of the two modes. Agreement was assessed using intraclass correlation coefficients (ICC), mean differences between electronic and paper scores, and Bland-Altman plots RESULTS: Forty-three patients, 12 male and 31 female, aged 32-83 years, completed the study. Overall agreements between the paper and electronic questionnaires was excellent (HAQ-DI: ICC = 0.963). Comparison of mean differences showed no evidence for bias in electronic vs. paper scores. The mean difference between electronic and paper scores was 0.03 units (S.D. 0.20), corresponding to 1% of the total scale length. Most patients found both paper and electronic questionnaires easy to use, and all patients found both modes acceptable. Six patients preferred paper, 23 preferred electronic, and 14 expressed no preference. CONCLUSIONS: There was excellent agreement between electronic and paper administration of the HAQ, suggesting that migration from paper to electronic had not led to any significant change in the data collected. Electronic methods of data collection were well-liked by patients and highly acceptable. The results support the use of the electronic version of this scale in clinical studies in rheumatoid arthritis