OBJECTIVE: Poor sleep quality is a commonly reported but under-investigated consequence of rheumatoid arthritis (RA). Actigraphy is a non-invasive way of measuring sleep, estimated from the frequency and intensity of physical movement at the wrist. We used actigraphy to measure sleep parameters compared with sleep questionnaire data, and assessed the practicality of actigraph use in patients with RA. METHODS: In a pilot study of actigraphy conducted within an investigation of night-time prednisone treatment and circadian interleukin-6 concentrations in ten patients with active RA, we compared actigraphy with the St Mary's Hospital Sleep Questionnaire and assessed whether night-time administration of prednisone resulted in increased sleep disturbance. RESULTS: The actigraph watch was well tolerated by our patients, producing adequate data for analysis for 128 out of 133 test days (96.2%). The results indicated reasonable concordance between actigraph and sleep questionnaire data in the present sample. Patient satisfaction with sleep (question 11) strongly correlated with sleep efficiency measured by the actigraph (r = 0.71, p = 0.22) and showed a trend for inverse correlation with the fragmentation index (r = -0.60, p = 0.067). Quality of sleep (question 9) correlated non-significantly with the fragmentation index (r = -0.59, p = 0.072). We were unable to identify any significant correlations between clinical measures of disease and sleep parameters in this sample. There were no apparent detrimental consequences of the night-time dose of prednisone on the measures of sleep quality and quantity. CONCLUSION: In spite of the physical disability imposed by RA, the actigraph was well tolerated and gave a useful measure of sleep in patients with active disease. It has the potential for use in larger controlled trials. Copyright © 2013 John Wiley & Sons, Ltd.