Pain sensitivity variations represent a critical frontier in pain neuroscience, where advanced neuroimaging has mapped cerebral correlates of nociception for decades, yet conventional brain-centric models persistently overlook the spinal cord's hub role in pain gating and amplification. Here we show that a corticospinal pain sensitivity signature, a pattern of functional connectivity from simultaneous corticospinal magnetic resonance imaging, predicts individual pain sensitivity and clinical pain. Trained on resting-state data and validated across independent healthy (n = 723) and patient cohorts (n = 46), the model generalized to new datasets, distinguished pain from non-pain, and outperformed brain-centric models. Crucially, transcranial magnetic stimulation perturbation revealed a causal axis where enhanced motor cortex-spinal connectivity directly changes pain perception (r = 0.55). These results indicate a previously unknown corticospinal biomarker that bridges laboratory pain measures and patient symptoms, providing insights into translating pain mechanisms from healthy individuals to clinical populations and informing neuromodulation approaches.
© 2025. The Author(s).