Motorneurons which are primarily affected in ALS cannot be sampled during life. Morphological, biochemical, ultrastructural and molecular investigations can be only performed on post-mortem material. No animal model reproduce adequately ALS. All these features and the low ALS incidence and the absence of definite clusters stress the need for brain banking in Europe. We present the protocol which is followed in our centers. The clinical information is supplied by neurologists in order to provide clinical data necessary for an accurate interpretation of pathological features. Initial symptomathology, duration and type of ALS, cause of death and a simple disability scale shortly before death are included. CNS relevant samples stained with H-E, luxol fast blue, the Marchi impregnation technique and immunostained with ubiquitin, are used for diagnosis. Samples for Golgi impregnation and immunohistochemistry are also obtained and the remaining tissue is frozen. Classical criteria for pathological diagnosis include: neurogenic changes in muscles, loss of motorneurons and degeneration of corticospinal tracts. We propose new diagnostic criteria based only on CNS examination. Muscle tissue is not always available and other classical criteria could be absent in rapid evolution or early death cases. Our criteria includes: a) Major criteria: loss or degeneration (chromatolysis, basophilia or neuronophagia) of motorneurons in brainstem and/or spinal cord, degeneration of corticospinal tracts. b) Minor criteria: axonal spheroids, Bunina bodies, ubiquitin-immunoreactives bodies. Criteria necessary for the pathological diagnosis are discussed and the need to quantify neuronal loss in relation to age-matched controls is stressed.