This study is an attempt to provide biomechanical support for the athletic practice of warming up prior to an exercise task to reduce the incidence of injury. Tears in isometrically preconditioned (stimulated before stretching) muscle were compared to tears in control (nonstimulated) muscle by examining four parameters: 1) force and 2) change of length required to tear the muscle, 3) site of failure, and 4) length-tension deformation. The tibialis anterior (TA), the extensor digitorum longus (EDL), and flexor digitorum longus (EDL) muscles from both hindlimbs of rabbits comprised our experimental model. Isometrically preconditioned TA (P less than 0.001), EDL (P less than 0.005), and FDL (P less than 0.01) muscles required more force to fail than their contralateral controls. Preconditioned TA (P less than 0.05), EDL (P less than 0.001), and FDL (P less than 0.01) muscles also stretched to a greater length from rest before failing than their nonpreconditioned controls. The site of failure in all of the muscles was the musculotendinous junction; thus, the site of failure was not altered by condition. The length-tension deformation curves for all three muscle types showed that in every case the preconditioned muscles attained a lesser force at each given increase in length before failure, showing a relative increase in elasticity, although only the EDL showed a statistically significant difference. From our data, it may be inferred that physiologic warming (isometric preconditioning) is of benefit in preventing muscular injury by increasing the and length to failure and elasticity of the muscle-tendon unit.