AIM: Spinal cord stimulation (SCS) is available as an alternative therapy for patients suffering from inoperable critical limb ischemia (CLI). Selection of patients is essential to achieve the best treatment effect. For this purpose transcutaneous oxygen (TcpO(2)) measurements have frequently been applied. So far, it is unclear which TcpO(2) parameters serve this purpose best. METHODS: Studies in which inoperable CLI patients were treated with conservative treatment with or without SCS, and in whom various TcpO(2) measurements were performed before and during treatment were pooled to investigate which TcpO(2) parameter(s) were best to detect patients who benefit most from SCS treatment as to limb salvage. RESULTS: TcpO(2) in the supine position increased significantly (p<0.001) in patients after a short period of SCS treatment (from 9 to 22 mmHg), but not in those treated conservatively (from 7 to 7 mmHg). Baseline supine TcpO(2) (using a cut-off value of 10 mmHg), the baseline sitting-supine TcpO(2) difference (cut-off value: 17 mmHg), and the difference in TcpO(2) before and after test stimulation (cut-off value: 4 mmHg) were related to a significantly increased limb salvage. SCS patients with a sitting-supine TcpO(2) difference of >17 mmHg had a 1-year limb salvage of 83% vs 68% in the whole SCS-treated group irrespective of TcpO(2) selection. CONCLUSION: The TcpO(2) parameters mentioned above are capable of detecting the effect of SCS treatment. Selection using (a combination of) TcpO(2) measurements substantially improves limb salvage of patients treated with SCS for inoperable CLI.