Many aspects of neurodegenerative disease pathology remain unresolved. Why do certain neuronal subpopulations acquire vulnerability to stress or mutations in ubiquitously expressed genes, while others remain resilient? Do these neurons harbor intrinsic marks that make them prone to degeneration? Do these diseases have a neurodevelopmental component? Lacking this fundamental knowledge hampers the discovery of efficacious treatments. While it is well established that human organoids enable the modeling of brain-related diseases, we still lack an organoid model that recapitulates the regionalization complexity and physiology of the spinal cord. Here, we describe an advanced experimental protocol to generate neuromuscular organoids composed of a wide rostro-caudal (RC) diversity of spinal motor neurons (spMNs) and mesodermal progenitor-derived muscle cells. This model therefore allows for the robust and reproducible study of neuromuscular unit development and disease.