Although, biomarkers are regarded as an important tool for monitoring injury severity and treatment efficacy, and for predicting clinical evolution in many neurological diseases and disorders including spinal cord injury, there is still a lack of reliable biomarkers for the assessment of clinical course and patient outcome. In this study, a biological dataset of 60 cytokines/chemokines, growth factorsm and intracellular and extracellular matrix proteins, analyzed in CSF within 24 h of injury, was used for correlation analysis with the clinical dataset of the same patients. A heat map was generated of positive and negative correlations between biomarkers and clinical rating scale scores at discharge, and between biomarkers and changes in clinical scores during the observation period. Using very stringent statistical criteria, we found 10 molecules which correlated with clinical scores at discharge, and five molecules, which correlated with changes in clinical scores. The proposed methodology may be useful for generating hypotheses regarding “predictive” and “treatment effectiveness” biomarkers, thereby suggesting potential candidates for disease-modifying therapies using a “bed-to-bench” approach.