Methods and systems are disclosed for evaluating a neurological condition by employing at least one motion sensor, configured to be attached to a body appendage, a memory associated with the sensor(s) to periodically record movement data during periods of prescribed exercises； and a processor for analyzing changes in movement data over time to evaluate the progression of the neurological condition. In one embodiment, the neurological condition is ALS and at least four motion sensors are employed such that each arm and leg of the patient has an associated sensor. The sensors can be accelerometers that measure the displacement, velocity and acceleration of an associated limb during periods of prescribed exercise. For example, changes in the patient's ability to repeat a series of limb-lifting exercises or the measurement of limb tremors associated with the conduct of the exercises can be correlated with norms and analyzed to classify the stage of ALS in a patient and/or predict the rate of progression.
