We have recently shown that subjects can learn to use cognitive-emotional strategies to suppress their spinal nociceptive flexor reflex (RIII reflex) under visual RIII feedback and proposed that this reflects learned activation of descending pain inhibition. Here, we investigated whether learned RIII suppression also affects supraspinal nociception and whether previous relaxation training increases success. Subjects were trained over 3 sessions to reduce their RIII size by self-selected cognitive-emotional strategies. Two groups received true RIII feedback (with or without previous relaxation training) and a sham group received false feedback (15 subjects per group). RIII reflexes, late somatosensory evoked potentials (SEPs), and F-waves were recorded and pain intensity ratings collected. Both true feedback groups achieved significant (P < 0.01) but similar RIII suppression (to 79% ± 21% and 70% ± 17% of control). Somatosensory evoked potential amplitude (100-150 milliseconds after stimulation) was reduced in parallel with the RIII size (r = 0.57, P < 0.01). In the sham group, neither RIII size nor SEP amplitude was significantly reduced during feedback training. Pain intensity was significantly reduced in all 3 groups and also correlated with RIII reduction (r = 0.44, P < 0.01). F-wave parameters were not affected during RIII suppression. The present results show that learned RIII suppression also affects supraspinal nociception as quantified by SEPs, although effects on pain ratings were less clear. Lower motor neuron excitability as quantified by F-waves was not affected. Previous relaxation training did not significantly improve RIII feedback training success.