We measured tibial translation in the knees of 22 sub jects with posterior cruciate ligament tears or recon structions by using the KT-1000 arthrometer. To as sess the reliability of the device, two testers made measurements. One tester was a novice and the sec ond tester had substantial clinical experience with the KT-1000 arthrometer. The quadriceps neutral angle was found for the uninvolved knee. Anterior and pos terior translation and total anterior-posterior excursion were then measured. The quadriceps neutral angle was then reproduced in the involved knee and the same measurements were taken. Each subject was tested twice by each tester. The intraclass correlation coefficient values for the novice, experienced, and in tertester reliability were 0.67, 0.79, and 0.62, respec tively, for corrected posterior translation; 0.59, 0.68, and 0.64, respectively, for corrected anterior transla tion; 0.70, 0.74, and 0.29, respectively, for quadriceps neutral angle; and 0.84, 0.83, and 0.62, respectively, for total anterior-posterior excursion. Ninety-five per cent confidence intervals for the novice, experienced, and intertester reliability were ±2.95, ±2.53, and ±3.27 mm, respectively, for corrected posterior trans lation; ±3.99, ±3.89, and ±3.74 mm, respectively, for corrected anterior translation; and ± 10.70°, ± 11.73°, and ± 16.25°, respectively, for quadriceps neutral an gle. The KT-1000 arthrometer was found to be a mod erately reliable tool for the measurement of tibial trans lation in patients with posterior cruciate ligament tears and reconstructions.