This study emanates from the desire to understand changes in neural signals observed when rats used a brain machine interface (BMI) to control a directional task. In this system, firing rates were estimated from several (8-10) motor cortical neurons at several consecutive time points. These firing rates were concatenated to form a neural activity vector (NAV). The NAV was used as input to a previously trained support vector machine (SVM). The decision function value obtained from the SVM was then used to determine which relay should be activated to replace paddle pressing signals for the task. Animals were able to do this immediately and significant changes in behavior and neural activity arose in a single, 45 minute, experimental session. Neural data from several subjects was examined for changes from the calibration phase to the late cortically controlled phase. These differences offer insight into how the rats and their SVMs collaborated to create a usable interface.