In recent decades, several novel approaches of spinal cord repair have revealed promising findings in animal models. However, for a successful translation of these into a clinical trial in humans the specific conditions pertaining to human spinal cord injuries (SCI) have to be appreciated. Firstly, transection of the spinal cord is commonly applied in animal models, whereas spinal cord contusion is the predominant type of injury in humans, and generally leads to more extensive injury in two to three spinal cord segments. Secondly, the quadrupedal organization of locomotion in animals and the more complex autonomic functions in humans challenge the translation of animal behavior into recovery from human SCI. Thirdly, so far, no adequate animal model has been developed to resemble spastic movement disorder in human SCI. Fourthly, the extensive damage to spinal motor neurons and nerve roots in human cervical and thoracolumbar in spine trauma is but little addressed in current translational studies. This damage has direct implications for rehabilitation and repair strategies. Fifthly, there is increasing evidence for a neuronal dysfunction below the level of the lesion in chronic complete SCI. The relevance of this dysfunction for a regeneration-inducing treatment needs to be investigated. Lastly, an approach to facilitate an appropriate reconnection of regenerating tract fibers by functional training in the postacute stage has yet to be confirmed.