Repair of defective knee ligaments requires knowledge of their normal function to achieve optimal surgical results and to prevent joint arthropathy. Many investigations have been conducted detailing the mechanics of the collateral ligaments of the human knee. A review of this literature, however, reveals a bewildering array of results, making it difficult to form confident conclusions concerning normal ligament function. The goal of this investigation was to clarify the past work with a systematic and direct investigation into the elongation behavior of the collateral ligaments, both passive and under external stress. The strain in the long, posterior parallel fibers of the medial collateral ligament (MCL) and in the middle third of the lateral collateral ligament (LCL) was tested under passive flexion/extension, varus and valgus rotations (3 degrees), internal and external tibial axial rotations (0 degrees-10 degrees), and quadriceps and hamstring loads (120 N). Each test was conducted over a flexion range of 15 degrees-120 degrees. Differential variable reluctance transducers (DVRTs), which are thought not to alter normal ligament function, were used to gather length change data. The transducers were calibrated against a linear variable differential transducer of a hydraulic testing machine (Model 858 Bionix, MTS System Corporation, MN), and output voltages were converted to length data in real time by a digital data acquisition system (TestStar II, Minneapolis, MN). A total of 12 knees were tested. Both the MCL and the LCL were shown to be more strained at extension than flexion. As would be expected, varus rotations increased the strain in the LCL and decreased the strain in the MCL, whereas valgus rotations increased the strain in the MCL and decreased the strain in the LCL. Internal tibial rotation was shown to decrease the strain in the MCL, while external tibial rotation increased the strain in the MCL. Interestingly, the LCL did not show consistent strain response to internal or external tibial rotation among cadavers, but was consistent between left and right knees from the same cadaver, indicating that the LCL is not a primary restraint to internal or external tibial rotation. Quadriceps loads decreased the strain in the MCL near extension and showed a trend of increased strain at extremes of flexion due to internal and external tibial rotations induced by the muscle load. The quadriceps loads created a highly consistent trend of strain reduction in the LCL for all flexion angles. The hamstring loads increased the strain in the MCL for large flexion angles (due to external tibial rotation created by hamstring loads) and non-significantly increased the strain in the LCL for all flexion angles. The reactions of the collateral ligaments to common external stresses must be known when evaluating injury mechanisms and attempting ligament repair. This study presented a complete picture of the strain behavior of the collateral ligaments as a tool for both the researcher and clinician.