This paper introduces a novel dataset designed to facilitate the understanding of medical prescriptions, addressing the critical need for accurate and efficient information extraction from semi-structured, sensitive patient health scanned documents. By structuring these prescriptions, medical professionals can more effectively verify dosage instructions and ensure the correct delivery of prescribed drugs to patients. Unlike existing document analysis datasets, Rx-PAD dataset is specifically tailored to the complex syntax and semantic relationships of medical prescriptions. Rx-PAD consists of 200 fully annotated images collected from French pharmacies, supporting two primary tasks: Entity Extraction (EE, 61 labels) and Entity Linking (EL, 11 relation types). To evaluate automatic prescription structuring, we propose a baseline model optimised for accurate detection of drugs and dosages. We also introduce the Drug Accuracy and Completeness (DAC) metric, which evaluates correct linking between drug names, dosings, and forms in real-world medical contexts. Rx-PAD provides a robust foundation for developing solutions that can handle sensitive medical data with the accuracy and speed required for real-world applications such as pharmacy automation, medical AI, and healthcare compliance applications, contributing to meaningful healthcare advancements. The dataset is publicly available at https://gitlab.com/phealing-public/rx_pad.