Literature findings on astrogliosis of the spinal cord and of the cortex of sporadic (SALS) and familial (FALS) cases of amyotrophic lateral sclerosis (ALS) and of SOD1 transgenic mice are analysed and compared with those of 50 autopsied personal cases of ALS. In the spinal cord, astrogliosis is definitely evident in the anterior horns and much less in anterior and lateral funiculi and dorsal horns. In the cortex reactive astrocytes show a distribution similar to that of ageing brains and in ischaemia and cannot be directly put in relation with neuronal loss. In the spinal cord of transgenic mice the evidence suggests a primary astrocytic response. The findings in human ALS, especially those of the cortex, are consistent with this hypothesis.