In order to assess the diagnostic accuracy of stress/rest myocardial perfusion scintigraphy in the follow-up of patients undergoing percutaneous transluminal coronary angioplasty (PTCA) we studied 50 patients (pts) before and 6 months after PTCA. All patients underwent control coronary angiography and then were divided in 2 groups, according to angiographic evidence of restenosis (25 pts) or no restenosis (25 pts). Myocardial perfusion imaging was performed with 99mTc-methoxy-isobutyl-isonitrile (MIBI). For MIBI scans, both qualitative and semi-quantitative analysis were performed. SPECT images were displayed on a color monitor in random order and graded blindly by 2 experienced observers. Rest and stress MIBI uptake was scored using a 4-point scale (ranging from 0 = normal, to 3 = absence of uptake). Individual subjects' perfusion scores were calculated by adding the individual segments' scores. History of relapsing angina showed a sensitivity and specificity of 76 and 96%, respectively. Exercise testing and MIBI tomoscintigraphy exhibited sensitivity and specificity of 80 and 56%, and 92 and 44%, respectively. However, when considering semiquantitative data and comparing them with pre-PTCA studies, specificity of MIBI scan increased to 96%, while sensitivity decreased to 72%. The results of the present study confirm high sensitivity and low specificity for both exercise ECG and myocardial perfusion scintigraphy, in the detection of restenosis following PTCA. However, when performing semiquantitative analysis of perfusion scanning and comparison with pre-PTCA images, the specificity of MIBI tomoscintigraphy increases significantly. Therefore, when adopting myocardial perfusion scintigraphy for the follow-up of patients undergoing PTCA, a pre-revascularization scan should be obtained as often as possible. This approach, in the context of a positive test, makes feasible the assessment of often partial improvements in perfusion of the myocardium relative to the treated vessel, consequently avoiding many false positive results which invariably lead to coronary angiography.