A locomotion analysis system for laboratory rats is presented. The system produces locomotion parameters (LPs) in 4 different domains: force, space, time and frequency. Video images of the walking rats are used to associate the system signals with individual limbs. Numerous LPs can be derived for every test run when the rat walks through the system on the way to sweets and a personal toy placed at the exit. This manuscript demonstrates that in order to differentiate SOD1-G93A mutant rat, a model of amyotrophic lateral sclerosis (ALS), from a Sprague Dawley (SD) control rat at a pre-symptomatic stage, one has only to use 8 key parameters. These 8 parameters are the bio-markers of ALS. The spline-based transformed values of these parameters are used as explanatory variables of a logistic regression model. This model predicts the probability that the examined rat belongs to the SOD1-G93A group. The model differentiates faultlessly between the SOD1 and control groups from the very first time the rats walked through the system at 51 days old. This system provides a new paradigm for ALS diagnosis, and it can have a significant impact on the development of new therapeutic procedures for ALS. The methodology presented in this manuscript can further address the development and validation of therapeutic procedures for other neurological diseases that affect locomotion.