OBJECTIVES: (1) To determine the proportion of patients with critical limb ischaemia refractory to invasive treatment that can be successfully treated by dorsal column stimulation (DCS), and (2) to identify potential responders to DCS by a simple test that is sufficiently predictive to limit the need for a prolonged trial stimulation period. PATIENTS AND METHODS: Twenty patients with chronic limb-threatening ischaemia were assessed by a battery of macro- and microcirculatory tests and a DCS trial of 1 week. Favourable response during the trial determined selection for long-term stimulation. RESULTS: During a mean observation period of 14 months, limb salvage rate (LSR) was 63% overall and 83% among patients selected after a favourable trial. Of the tests performed preoperatively, digital subtraction angiography, Doppler assessment, oscillometry, capillaroscopy, foot temperature, and transcutaneous partial pressure of oxygen (tcpO2) in the supine or sitting position did not adequately predict DCS response. The supine-sitting tcpO2 gradient (delta tcpO2) was a good predictor of DCS outcome, with an 88% LSR when delta tcpO2 > 15 Torr, dropping to 12% when delta tcpO2 < or = 15 Torr. CONCLUSIONS: DCS is a rewarding therapeutic option in selected patients with critical limb ischaemia. delta tcpO2 appears to reliably predict response to DCS and may obviate trial stimulation in most cases.