STUDY DESIGN: Experimental, prospective, blinded, animal study. OBJECTIVES: Subtotal transection models in rodents are widely used in spinal cord injury (SCI) research. In this model, we investigate the effect of the dorso-ventral incision depth (ID) of the spinal cord on functional locomotor outcome using the Basso, Beattie, Bresnahan (BBB) scale. We introduce the minimal incision depth (ID(min)) and the average lesion depth (ID(mean)) as reliable, fast and easily available predictive parameters for intrinsic locomotor function. SETTING: Tuebingen, Germany. METHODS: Dorsal over-hemisection at the level of T8 was performed in male Lewis rats. Functional outcome 4 weeks after SCI and histological analysis of the lesion were studied and correlated in 36 animals. Animals reaching weight support (BBB> or =9) were considered as having reached functional recovery. Data analysis was performed in linear (ordinary least squares; OLS) and nonlinear (logistic) regression models for correlation of histological parameters and functional outcome. RESULTS: BBB scores revealed a strong correlation with ID(mean) and ID(min), showing a higher value in predicting functional outcome for the latter parameter. Based on logistic regression analysis, animals with an ID(min) of 69% would have a 95% probability of reaching weight support. CONCLUSION: These results demonstrate that histological analysis is crucial when functional outcome parameters are used in the dorsal over-hemisection SCI model. A simple and feasible histological evaluation can reliably predict spontaneous functional locomotor recovery in dorsal transection models and could provide a simple tool to identify treatment effects of new experimental therapeutic approaches.