We adapted the method of 99mTc-pertechnetate (99mTc) uptake measurements to the mouse knee for detection and quantification of arthritis because clinical assessment of mouse knee-joint arthritis is not reliable. The main points to ensure reproducibility of measurements were proper fixation and positioning of the knee and careful shielding of the rest of the body from the gamma-radiation detector. 99mTc uptake was calculated as the mean of three countings. The variation coefficient of these countings ranged from 0.007 to 0.082 in non-arthritic joints and from 0.025 to 0.081 in arthritic joints. Arthritis was scored as the ratio of the 99mTc uptake in the right knee versus that in the left knee. This ratio averaged 1.06 (S.D. 0.05) in non-arthritic mice 20 minutes after 99mTc administration i.p. On the second day after induction of arthritis in the right knee, this ratio ranged from 1.44 to 1.96; this was significantly higher (P less than 0.005) than in non-arthritic mice, and the increase remained significant during at least 20 days. 99mTc uptake measurements seem to be a useful method to detect and quantify arthritis of the knee joint in mice.