Adeno-associated viral (AAV) vectors are a platform of great potential for therapeutic gene delivery. One of the major challenges regarding AAV gene therapy is to deliver the transgene of interest to target cells at levels that result in expression that is both safe and effective. For diseases of the central nervous system (CNS) such as amyotrophic lateral sclerosis (ALS) and Friedreich's ataxia, it is important to identify a dosing paradigm that provides a relatively homogenous distribution of gene transfer along the rostral-caudal axis of the spinal column in the CNS and is translatable. Intrathecal (IT) administration is a delivery approach that has shown promise for providing such distribution. AAV dosing via the IT route has been reported in large mammals to provide greater CNS distribution, less exposure to peripheral organs and tissues, and reduced impact of immune responses than systemic dosing. However, to-date, IT dosing of AAV in large mammals has been investigated primarily by lumbar bolus administration, with a few studies assessing administration at more rostral sites either alone or in combination with a lumbar site. Parameters that are likely to have a major impact on CNS distribution such as volume, rate and duration of infusion have not been reported previously for IT dosing of AAV. Here, we studied the effects of these parameters as well as site of infusion on distribution of transgene expression in the non-human primate CNS, using AAVrh10 to package a vector genome (vg) containing the human frataxin gene driven by the chicken β-actin promoter. Four weeks after dosing, frataxin (FXN) expression was assessed in spinal cord and dorsal root ganglia (DRG) at multiple rostral-caudal levels, as well as in brain, cerebellum, and peripheral organs such as liver, spleen and heart. Quantitative assessments included measurements of vector genome copy number and mRNA and protein expression levels. Our results showed that the distribution of transgene expression in the spinal cord, DRG, cerebellum and brain were significantly impacted by the site(s) of IT delivery (i.e. cervical vs lumbar), and volume, rate and duration of infusion. As expected, there was transduction of peripheral organs, indicating systemic exposure of the AAV vector following IT delivery. In addition, FXN expression across different cell types is being studied by histological methods. These results will not only guide the optimization of IT delivery of AAV gene therapy for diseases such as ALS and Friedreich's ataxia, but also provide useful information for IT dosing for other CNS disorders.