The biomechanical response of the in vitro lumbar motion segment (functional spinal unit, FSU) under a dynamic (transient) flexion-shear load was determined. The load was transferred to the specimen by a padded pendulum and simulated a flexion-distraction injury, a so called lap seat-belt injury. The load response of the specimen was measured with a force and moment transducer, and the motions were determined with high speed photography. Two series of tests were made with 10 specimens in each, with two different load pulses: one moderate pulse (mean acceleration 2.5 g. duration 150 ms) and one severe pulse (mean acceleration 8 g, duration 250 ms). The results showed that the moderate load pulse caused initial flexiondisuaction injuries at a mean bending moment of 113 Nm and a mean shear force of 346 N. The maximum flexion angulation attained during the loading sequence was 14’. The severe load pulse caused evident signs of failure or total rupture of the segments at a mean bending moment of 151 Nm and a mean shear force of 481 N. The flexion angulation just before failure was 19’. A statistically significant correlation (r>O.7, p<0.05) vias found between the load response and the height of the segment, the load response and the lateral disc diameter, and the load response and the bone mineral content (BMC) in the vertebrae. Comparisons were made with previous established thresholds for static flexion-shear loading. The results indicated that thresholds for initial and ultimate flexion-disuaction injury respectively are in the same range for static and uansient loading conditions.