Neurodegenerative diseases have been construed as incurable disorders. However, therapeutic development for these diseases is now facing a turning point: analyses of cellular and animal models have provided insights into pathogenesis of neurodegenerative diseases, and have indicated rational therapeutic approaches to them. Therefore, how to design clinical trials for neurodegenerative diseases is becoming one of the most challenging issues in the clinical neurology. Given a small number of patients and a slow progression of symptoms, it is important to carry out effective trials using surrogate endpoints or historical control data, if necessary. In particular, pathogenesis-based disease-modifying therapy should be carefully evaluated in clinical trials, because its effect is likely to be suppression of neurodegenerative processes, but not symptom relief. The development of androgen deprivation therapy for spinal and bulbar muscular atrophy (SBMA) is a representative study in this field. SBMA is a hereditary neurodegenerative disease caused by expansion of a trinucleotide CAG repeat, which encodes the polyglutamine tract, in the first exon of the androgen receptor (AR) gene. Animal models carrying human mutant AR gene recapitulate polyglutamine-mediated motor neuron degeneration, providing insights into the pathogenesis of SBMA. There is increasing evidence that testosterone, the ligand of AR, plays a pivotal role in the neurodegeneration in SBMA. The striking success of androgen deprivation therapy in SBMA mouse models has been translated into phase 2, and then phase 3, clinical trials.