Standard clinical methods of assessing joint inflammation are being supplemented increasingly by radioisotopic and thermographic studies. However, the correlation between these different methods has not been firmly established. In the quantification of synovitis by infrared thermography we have shown that the heat distribution index (HDI) based on thermal pattern is more reliable and is less affected by diurnal variations in joint temperature than the commonly used thermographic index, which is based on average skin temperature values. In 20 patients with rheumatoid arthritis whose knees were being treated with intra-articular steroid we obtained 184 serial paired observations over a period of 24 weeks for clinical assessment, HDI, and 99mTc pertechnetate uptake. We found significant correlations (p less than 0.001) between the three methods of assessment (except for pain and HDI (p = 0.116)).