Translation of pharmacological results from in vitro cell testing to clinical trials is challenging. One of the causes that may underlie these discrepant results is the lack of the phenotypic or species-specific relevance of the tested cells; today, this lack of relevance may be reduced by relying on cells differentiated from human pluripotent stem cells. To analyse the benefits provided by this approach, we chose to focus on Friedreich ataxia, a neurodegenerative condition for which the recent clinical testing of two compounds was not successful. These compounds, namely, resveratrol and nicotinamide, were selected because they had been shown to stimulate the expression of frataxin in fibroblasts and lymphoblastoid cells. Our results indicated that these compounds failed to do so in iPSC-derived neurons generated from two patients with Friedreich ataxia. By comparing the effects of both molecules on different cell types that may be considered to be non-relevant for the disease, such as fibroblasts, or more relevant to the disease, such as neurons differentiated from iPSCs, a differential response was observed; this response suggests the importance of developing more predictive in vitro systems for drug discovery. Our results demonstrate the value of utilizing human iPSCs early in drug discovery to improve translational predictability.