Amyotrophic lateral sclerosis (ALS) is characterized by adult-onset progressive degeneration of upper and lower motor neurons. Increasing numbers of genes are found to be associated with ALS; among those, the first identified gene, SOD1 coding a Cu/Zn-superoxide dismutase protein (SOD1), has been regarded as the gold standard in the research on a pathomechanism of ALS. Abnormal accumulation of misfolded SOD1 in affected spinal motor neurons has been established as a pathological hallmark of ALS caused by mutations in SOD1 (SOD1-ALS). Nonetheless, involvement of wild-type SOD1 remains quite controversial in the pathology of ALS with no SOD1 mutations (non-SOD1 ALS), which occupies more than 90% of total ALS cases. In vitro studies have revealed post-translationally controlled misfolding and aggregation of wild-type as well as of mutant SOD1 proteins; therefore, SOD1 proteins could be a therapeutic target not only in SOD1-ALS but also in more prevailing cases, non-SOD1 ALS. In order to search for evidence on misfolding and aggregation of wild-type SOD1 in vivo, we reviewed pathological studies using mouse models and patients and then summarized arguments for and against possible involvement of wild-type SOD1 in non-SOD1 ALS as well as in SOD1-ALS.