The low regeneration potential of the central nervous system (CNS) represents a challenge for the development of new therapeutic strategies for neurodegenerative diseases, including spinocerebellar ataxias. Spinocerebellar ataxia type 3 (SCA3)—or Machado–Joseph disease (MJD)—is the most common dominant ataxia, being mainly characterized by motor deficits; however, SCA3/MJD has a complex and heterogeneous pathophysiology, involving many CNS brain regions, contributing to the lack of effective therapies. Mesenchymal stem cells (MSCs) have been proposed as a potential therapeutic tool for CNS disorders. Beyond their differentiation potential, MSCs secrete a broad range of neuroregulatory factors that can promote relevant neuroprotective and immunomodulatory actions in different pathophysiological contexts. The objective of this work was to study the effects of (1) human MSC transplantation and (2) human MSC secretome (CM) administration on disease progression in vivo, using the CMVMJD135 mouse model of SCA3/MJD. Our results showed that a single CM administration was more beneficial than MSC transplantation—particularly in the cerebellum and basal ganglia—while no motor improvement was observed when these cell-based therapeutic approaches were applied in the spinal cord. However, the effects observed were mild and transient, suggesting that continuous or repeated administration would be needed, which should be further tested.