Previous studies have used various techniques for microscopic analysis of rheumatoid synovium, ranging from rapid analysis of limited areas of tissue to detailed quantification of extensive areas. The sensitivity and reproducibility of these methods have not been tested. This study sought to determine the minimum area of rheumatoid synovium needed to allow accurate microscopic analysis of synovial inflammation. Multiple synovial tissue samples were obtained from patients with rheumatoid arthritis at knee arthroplasty (n = 10), knee arthroscopy (n = 10) and by blind needle biopsy (n = 23). Lining layer thickness, sublining T-cell infiltration and vascularity were measured in all high-power fields (hpf) throughout every sample obtained from each patient. These complete measured results were compared with estimated results from limited numbers of hpf from each patient. It was observed that lining layer thickness estimated from as few as five readings from 3 samples/patient correlated significantly with the measured results obtained from as many as 85 readings/patient [Tau (T) = 0.70-0.94 for the three groups, all P < or = 0.005). Estimated measures of T-cell infiltration and vascularity derived from only 17 randomly selected hpf from 3 samples/patient (equivalent to 1 mm2) correlated significantly with the measured results obtained from up to 150 hpf/patient (T = 0.65-0.94, all P < or = 0.002). Quantitative analysis of inflammation in synovial tissue samples is both accurate and practical when restricted to an evaluation of a limited number of microscopic fields. It is proposed that lining layer thickness may be confidently quantified from five randomly selected readings from three tissue samples, and that sublining T-cell infiltration and vascularity may be quantified from 17 randomly selected hpf from the same samples.