Central pain below the injury level after spinal cord injury is excruciating, chronic and resistive to treatment. Animal studies suggest that pretreatment may prevent central pain, but to date there are no measures to predict its development. Our aim was to monitor changes in the sensory profile below the lesion prior to the development of below-level central pain in order to search for a parameter that could predict its risk and to further explore its pathophysiology. Thirty patients with spinal cord injury and 27 healthy controls underwent measurement of warm, cold, heat-pain and touch thresholds as well as graphaesthesia, allodynia, hyperpathia and wind-up pain in intact region and in the shin and feet (below level). Patients were tested at 2-4 weeks, 1-2.5 months and 2.5-6 months after the injury or until central pain had developed. At the end of the follow-up, 46% of patients developed below-level central pain. During the testing periods, individuals who eventually developed central pain had higher thermal thresholds than those who did not and displayed high rates of abnormal sensations (allodynia and hyperpathia), which gradually increased with time until central pain developed. Logistic regressions revealed that the best predictor for the risk of below-level central pain was allodynia in the foot in the second testing session with a 77% probability (90.9% confidence). The results suggest that neuronal hyperexcitability, which may develop consequent to damage to spinothalamic tracts, precedes central pain. Furthermore, it appears that below-level central pain develops after a substantial build-up of hyperexcitability. To the best of our knowledge, this is the first systematic report establishing that neuronal hyperexcitability precedes central pain. Predicting the risk for central pain can be utilized to initiate early treatment in order to prevent its development.