To obtain optimal image quality in myocardial viability studies, it is recommended that 18F-fluordeoxyglucose (18F-FDG) studies be performed with hyperinsulinaemic glucose clamping. 18F-FDG imaging after oral administration of acipimox, a nicotinic acid derivative, results in comparable image quality to clamping. Twenty consecutive patients (7 with diabetes mellitus) with angiographically confirmed coronary artery disease and similar demographic/clinical profiles were randomly allocated to gated cardiac 18F-FDG-PET with a standard euglycaemic hyperinsulinaemic clamp protocol or using a combination of oral administration of acipimox and the insulin clamp technique. The image quality, expressed as the myocardial-to-blood pool activity ratio, was superior in the combined protocol compared with the insulin clamping technique alone (3.37 +/- 1.46 vs 2.27 +/- 0.62, P = 0.037). Although there were no significant differences in plasma insulin and free fatty acids concentrations between the two protocols, plasma glucose concentrations obtained with the standard protocol were elevated compared with the combined protocol (11.1 +/- 3.7 vs 6.3 +/- 3.0 mM during clamping; 10.2 +/- 3.3 vs 5.5 +/- 3.0 mM during acquisition). We conclude that gated 18F-FDG-PET imaging after oral administration of acipimox plus insulin clamping yields image quality superior to that obtained with clamping alone.