We evaluated the accuracy of six clinical tests for posterior instability in 24 knees with acute surgically-proven posterior cruciate ligament injuries and intact anterior cruciate ligaments. We also performed stress radiography under anaesthesia. The gravity sign and the posterior drawer test in near extension and its passive reduction were diagnostic in 20 of the 24 knees, and the active reduction of posterior subluxation was diagnostic in 18. The reversed pivot shift sign helped to diagnose severe posterior and posterolateral subluxations, but the external rotation recurvatum test was negative in all 24 knees. Stress radiography in near extension revealed a highly significant increase in posterior tibial subluxation in the injured knees.

A logical, objective and reproducible grading system for the pivot shift test is proposed. The rationale is based on performing the examination in varying positions of rotation of the tibia, allowing the type and degree of the different laxities to be defined and quantified. The system has been assessed against a new "unblocked" test for anterior subluxation and against radiographic measurements, operative findings and results. This grading system can be valuable in pre-operative assessment and planning and its use in postoperative evaluation would enable results from different centres and different procedures to be compared more accurately.