Background. We assessed mirror visual feedback (MVF) to test the hypothesis that incongruence between motor output and sensory input produces complex regional pain syndrome (CRPS) (type 1) pain. Methods. Eight subjects (disease duration greater than or equal to3 weeks to less than or equal to3 yr) were studied over 6 weeks with assessments including two controls (no device and viewing a non-reflective surface) and the intervention (MVF). Pain severity and vasomotor changes were recorded. Results. The control stages had no analgesic effect. MVF in early CRPS (less than or equal to8. weeks) had an immediate analgesic effect and in intermediate disease (less than or equal to1 yr) led to a reduction in stiffness. At 6 weeks, normalization of function and thermal differences had occurred (early and intermediate disease). No change was found in chronic CRPS. Conclusions. In early CRPS (type 1), visual input from a moving, unaffected limb re-establishes the pain-free relationship between sensory feedback and motor execution. Trophic changes and a less plastic neural pathway preclude this in chronic disease.