PURPOSE OF THE STUDY To improve the important torsional, bending and compressive stability in femoral neck fixation, locking plates have been the latest contribution. However, increased strength by restricted fracture motion may come at expense of an altered load distribution and failure patterns. Within locking plate technology, the important intermediate fracture compression may principally be achieved by multiple sliding screws passing through a sideplate fixed to the femur or connected to an interlocking plate not fixed to the femur laterally, sliding "en bloc" with the plate. While biomechanical studies may deliver the short-time patient safety requirements in implant development, no adequate failure evaluation has been performed with interlocking devices ex vivo in this setting. In the present biomechanical study, we analysed if a novel femoral neck interlocking plate with pins could improve fixation performance by changing the parameters involved in the failure mechanism in terms of fixation strength, fracture motion, load distribution and failure pattern. MATERIAL AND METHODS Sixteen pairs of human femurs with stable subcapital osteotomies were fixated by 2 pins or 3 pins interlocked in a plate using a paired design. Femurs were loaded non-destructively to 10° torsion around the neck axis, 200 N anteroposterior bending and 500 N vertical compression in 7° adduction with 1 Hz in 20 000 cycles, and were subsequently subjected to destructive compression to evaluate failure patterns. Bending stiffness, compressive stiffness and displacement from compressive testing reflected fracture motion. Torque and compression to failure replicated known failure mechanisms and defined strength. To evaluate load distribution, associations between biomechanical parameters and measured local bone mineral measurements by quantitative CT were analysed. RESULTS Interlocked pins increased mean strength 73% in torsion and 39% in compression (p = 0.038). Strength was related to all 4 regional mineral masses from the femoral head to subtrochanterically with interlocking (r = 0.64-0.83, p = 0.034), while only to mineral masses in the femoral head in compression and to the head, neck and trochanterically in torsion with individual pins (r = 0.67-0.78, p = 0.024). No difference was detected in fracture motion or failure pattern. DISCUSSION Within the last decade, angular stable implants have expanded our therapeutic arsenal of femoral neck fractures. Increased stability at the expense of altered devastating failure patterns was not retrieved in our study. The broadened understanding of the effect of interlocking pins by an isolated plate as in the current study involved the feature to gain fixation strength. By permitting fracture compression, and through a significant change of correlations between mechanical parameters and local bone mineral factors, a lateral redistribution of load with interlocked pins from the fragile bone medially to the more solid lateral bone was demonstrated. Regarding the long-term patient safety of interlocked pins and healing complications of non-union and segmental collapse of the femoral head, a definite conclusion may be premature. However, the improved biomechanics of an interlocking plate must be considered a favourable development of the pin concept. CONCLUSIONS Interlocked pins may improve fixation performance by a better load distribution, not by restricting fracture motion with corresponding altered failure patterns. This is encouraging and a challenge to complete further studies of the interlocking plate technology in the struggle to find the optimal treatment of the femoral neck fracture. Key words: femoral neck fracture, biomechanics, cadaver bone, bone mineral, internal fixation, locking plate, interlocked pins.