The purpose of this study was to validate the measurement of global and regional left ventricular cardiac function with ECG-gated fluorine-18 fluorodeoxyglucose (FDG) positron emission tomography (PET) by comparison with the corresponding indices from X-ray left ventriculography (LVG). Twenty-six patients (23 men, 3 women, mean age 60.4 years) underwent LVG and ECG-gated (eight frames/cycle) FDG-PET within an interval of 10.2+/-6.8 days. A volumetric sampling approach was used to obtain both global (EF: ejection fraction) and regional [%WT: relative regional count increase from end-diastolic (ED) to end-systolic (ES) phase] functional parameters. The gated PET parameters were compared with the corresponding findings of LVG in seven myocardial segments. EF(gated PET) and EF(LVG) did not differ significantly (30%+/-10% vs 32%+/-10%, P=NS). The two EF values correlated significantly, showing no significant systematic measurement bias [EF(gated PET) = 2.61+0.86 x EF(LVG), R=0.84, P<0.0001]. Inter- and intra-observer reproducibility for EF were R=0.95, P<0.0001 and R=0.92, P<0.0001, respectively. Regional function was evaluated with LVG in 144 myocardial segments comprising 35 normokinetic, 70 hypokinetic and 39 a/dyskinetic segments. Visual analysis of LVG and gated PET correlated significantly (P<0.001), with an overall concordance ratio of 58% (83/144, kappa=0.35). Gated PET overestimated the regional function in 27% (39/144) and underestimated it in 15% (22/144). %WT showed significant differences between each pair of groups (a/dyskinesis, 13.2%+/-9.3%; hypokinesis, 17.1%+/-8.8%; normokinesis, 21.8%+/-10.9%). Inter- and intra-observer reproducibility was significant for %WT (R=0.77, P<0.0001 and R=0.79, P<0.0001, respectively). In conclusion, gated FDG PET permits assessment of global left ventricular cardiac function. In addition, assessment of regional function is feasible using the visual or the quantitative parameters.