Spinal cord stimulation (SCS) can relieve symptoms in patients with severe angina pectoris refractory to conventional medical or surgical therapy. This symptomatic improvement may result from decreased myocardial ischemia. To test this hypothesis, positron emission tomography (PET) and potassium-38 as a flow tracer were used in 8 patients for the quantitative evaluation of regional myocardial perfusion at rest and after exercise, before and during SCS. Potassium uptake was evaluated as myocardial clearance (flow times net extraction) in ml/min/100 g. Tomographic segments were categorized as nonaffected and affected on the basis of the absence or presence of arterial stenosis on coronary angiography and on the basis of thallium scintigraphic data. In nonaffected segments, before SCS, regional myocardial clearance significantly increased from rest (28 +/- 4) to exercise (47 +/- 13 clearance units; p less than 0.004). A similar increase occurred after SCS. In affected segments, before SCS, regional myocardial clearance barely increased (p = 0.065) from rest (26 +/- 6) to exercise (33 less than or equal to 12). In comparison, after SCS, the resting regional myocardial clearance was slightly elevated (29 +/- 8) reflecting an increased double product, but did not increase (p = 0.192) with exercise (34 +/- 12). However, the magnitude and duration of ST-segment depression decreased during treatment with SCS. Anginal pain occurred in all patients during control exercise, but was attenuated in all but one with SCS. These results indicate that SCS improves exercise-induced angina and electrocardiographic signs of ischemia but this influence does not appear to be mediated by changes in regional myocardial perfusion.