This comparative study was performed to determine whether a conventional planar gamma camera optimised for 511-keV imaging can reliably assess myocardial viability using the fluorine-18 fluorodeoxyglucose (FDG) metabolic tracer previously developed for positron emission tomography (PET). Twenty-seven patients with severe ischaemic cardiomyopathy (mean left ventricular ejection fraction: 20% +/- 9%) having clinically indicated nitrogen-13 ammonia/FDG PET myocardial viability studies consented to resting, four-view, planar myocardial thallium-201 perfusion and FDG metabolism imaging. The resultant PET and planar perfusion/metabolism images (PPI) were independently assessed for FDG defect size and perfusion/metabolism mismatch, using a four-point scale, in each of four vascular regions: apex, circumflex, left anterior and posterior descending coronary artery territories. Of 108 regions, 106 were evaluable (two not assessed by PET). There was complete agreement in 70% of coronary vascular territories, giving an unweighted kappa score of 0.56. Moreover, in 94% of segments agreement was within one grade. Interestingly, six of the seven differences of more than one grade occurred in the circumflex coronary territory, which was also the only region for which planar positron imaging underestimated FDG defect size. Three of four moderate areas of perfusion/metabolism mismatch seen with PET were also seen on PPI. PPI showed three small regions of mismatch not seen on PET, whilst the reverse occurred with one other small region of mismatch. Thus, for this PET protocol, PPI provides very similar information on the extent of regional FDG uptake and occurrence of mismatch.(ABSTRACT TRUNCATED AT 250 WORDS)