Unreliable mouse models have been the downfall of many promising therapies for neurological disease. Here we conduct safety and efficacy trials of an AAV9 gene therapy for hereditary spastic paraplegia 56 (SPG56) in patient-derived brain organoids to more accurately recapitulate this neurological disorder and with the added objective of determining disease mechanisms. SPG56 is characterised by early onset spasticity and axonal neuropathy and is caused by mutations in CYP2U1, a CYP450 enzyme involved in long chain fatty acid metabolism. We generated iPSCs from two SPG56 patients and their heterozygous parents, and used these lines to generate neuronal and brain organoid cultures. We subsequently treated brain organoids with EF1A-CYP2U1-AAV9 and measured CYP2U1 enzymatic activity, mitochondrial function, and conducted transcriptomic, proteomic and metabolomic analysis. Mouse trials were performed in parallel to explore the potential of organoids as a pre-clinical platform. We found that CYP2U1^mut organoids and neuron cultures had impaired neuronal morphology and electrical function, reduced energy generation (altered Kreb cycle metabolites and down regulated fatty acid β-oxidation pathways), and impaired mitochondrial properties, including respiration. Treatment with AAV9 gene therapy increased CYP2U1 expression, restored neuronal processes and significantly altered levels of Krebs cycle metabolites and transcriptomic pathways. By comparing brain organoid data to data obtained from CYP2U1-AAV9 mouse trials conducted in parallel, we not only provide the data required to progress to clinical trials in a small number of patients, but are also validating the use of brain organoids for pre-clinical gene therapy.