The glycosphingolipid (GSL) lysosomal storage diseases are a family of human metabolic diseases that, in their severest forms, cause death in early infancy, as a result of progressive neurodegeneration. They are caused by mutations in the genes encoding the glycohydrolases or the activator proteins that catabolise GSLs within lysosomes. In these diseases the GSL substrate of the defective enzyme accumulates in the lysosome, where it is stored and leads to cellular dysfunction and disease. The therapeutic options for treating these diseases are relatively limited; in fact, there are currently no available therapies for most of these disorders. The problem is further compounded by difficulties in delivering therapeutic agents to the central nervous system, which is where the pathology is frequently manifested. To date, research effort has mainly focused on strategies for augmenting enzyme concentrations to compensate for the underlying defect. These strategies include bone-marrow transplantation, enzyme-replacement therapy and gene therapy. Our group has been exploring the alternative strategy of substrate deprivation. This approach aims to balance the rate of GSL synthesis with the impaired rate of GSL breakdown. Studies using an asymptomatic mouse model of Tay–Sachs disease have shown that substrate deprivation prevents GSL storage. In a severe neurodegenerative mouse model of Sandhoff disease, substrate deprivation delayed the onset of symptoms and disease progression, and significantly increased life expectancy. The implications of this research for human therapy have been discussed.