BACKGROUND: Descending pain inhibition suppresses spinal nociception, reducing nociceptive input to the brain. It is modulated by cognitive and emotional processes. In subjects with chronic pain, it is impaired, possibly contributing to pain persistence. A previously developed feedback method trains subjects to activate their descending inhibition. Participants are trained to use cognitive-emotional strategies to reduce their spinal nociception, as quantified by the nociceptive flexor reflex (RIII reflex), under visual feedback about their RIII reflex size. The aim of the present study was to test whether also subjects with chronic back pain can achieve a modulation of their descending pain inhibition under RIII feedback. METHODS: In total, 33 subjects with chronic back pain received either true (n = 18) or sham RIII feedback (n = 15), 15 healthy control subjects received true RIII feedback. RESULTS: All three groups achieved significant RIII suppression, largest in controls (to 76 ± 26% of baseline), intermediate in chronic back pain subjects receiving true feedback (to 82 ± 13%) and smallest in chronic back pain subjects receiving sham feedback (to 89 ± 14%, all p < 0.05). However, only chronic pain subjects receiving true feedback significantly improved their descending inhibition over the feedback training, quantified by the conditioned pain modulation effect (test pain reduction of baseline before training: to 98 ± 26%, after: to 80 ± 21%, p < 0.01). CONCLUSION: Our results show that subjects with chronic back pain can achieve a reduction of their spinal nociception and improve their descending pain inhibition under RIII feedback training. SIGNIFICANCE: Subjects with chronic back pain can learn to control their spinal nociception, quantified by the RIII reflex, when they receive feedback about the RIII reflex.