The aim of this study was to directly compare three currently used quantitative methods of analysis of technetium-99m sestamibi images in patients with selective balloon-induced transmural ischaemia. The area at risk (AR) was assessed in 19 patients undergoing single-vessel percutaneous transluminal coronary angioplasty by injecting the 99mTc-sestamibi at the time of coronary artery occlusion during balloon inflation. After imaging, the patients were classified according to localization of the perfusion defect as having anteroseptal (group I, 11 patients) or posterolateral defects (group II, eight patients). The planimetric technique based on polar maps, proposed by Verani et al. (J Am Coll Cardiol, 1988) (method A), the method described by Tamaki et al. (Circulation, 1982) (method B) and the technique validated by O Connor et al. (Eur J Nucl Med, 1990) (method C) were tested. Three threshold values of 45%, 50% and 60% of the maximum left ventricular count were used to define the limits of the perfusion defect. The mean values of the AR calculated by the three techniques with the original cut-off level (method A=16.5%+/-12.9; method B=10.4%+/-7.6%; method C=29.6%+/-15.7%) were statistically different (one-way analysis of variance: P<0.001; paired t test: method A vs B, P=0.003; method B vs C and method A vs C, P<0.0001). There was no significant difference between the mean values of the AR estimated by the three methods using the same cut-off levels. The use of 60% of the maximum left ventricular count provided the best correlation between the techniques (method A vs B, r=0.95; method B vs C, r=0.92; method A vs C, r=0.95). Nevertheless, a difference >10% between the values of AR assessed by the three methods was found in four subjects. There was no significant difference between the three methods in the evaluation of AR in the subjects of group I and group II. Reproducibility was good for all methods. It is concluded that the three methods of analysis of the AR by 99mTc-sestamibi SPET imaging showed comparable performance and good reproducibility using the same cut-off level. The location of perfusion defect does not affect the comparability of the three techniques. We suggest the use of a cut-off level of 60% for all three methods in the assessment of the AR by 99mTc-sestamibi SPET imaging.