The spinal muscular atrophies (SMAs) are defined as a group of inherited disorders sharing the common pathological feature of degeneration of the anterior horn cells of the spinal cord and, in some cases, additionally of the bulbar motor nuclei. They are classified according to clinical features, including severity and distribution of muscle weakness and on their modes of inheritance. The SMAs are not uncommon: SMA type I (severe, acute infantile SMA) alone has a gene frequency in the UK estimated at 0.006. Together they represent a significant source of morbidity and mortality. Over the past 3 years, two major advances have been made towards understanding the molecular basis of these clinically heterogeneous disorders. First, in 1990, it was reported that all three forms of childhood-onset proximal SMA were linked to probes mapping to the proximal long arm of chromosome 5. Significant progress has been made towards isolating the gene or genes responsible, and a protocol for prenatal disease prediction in families with a previously severely affected child has been established. Second, in 1991, the molecular defect in adult-onset X-linked spinal and bulbar muscular atrophy was defined as an expansion of a (CAG)n trinucleotide repeat which encodes a string of glutamine residues in the first exon of the androgen receptor. Little progress has been made in elucidating the molecular pathology of the other SMAs. None of them has been linked to chromosome 5q markers, or indeed to markers elsewhere; the conditions are rare and pedigrees generally small. When the gene (or genes) on proximal 5q underlying SMA types I, II and III is cloned, mutations in this can then be sought in the other SMAs. The product of the chromosome 5q gene presumably plays a crucial role in maintaining the integrity of motor neurones, so determination of its structure and function may well have consequences far beyond those pertaining to the inherited SMAs. The identification of pathogenic mutations in the SOD-1 gene in familial amyotrophic lateral sclerosis is of great interest, although it is not clear that similar mechanisms contribute to the more common sporadic form of the disease.