This study was conducted to evaluate the frequency and possible causes of poor red blood cell (RBC) labelling when performing equilibrium gated blood pool (GBP) radionuclide angiography at rest with an in vivo method. The influence of the mode of administration on tagging efficiency was studied by investigating the image quality in 160 patients referred for evaluation of left ventricular (LV) function prior to or after coronary angiography, while using a roughly standardized administration protocol. The patients were subclassified into four groups according to the way both molecules involved in the tagging procedure were administered. When poor image quality was found (in 9.4% of the patients), the labelling efficiency was quantified and the frequency of failed tagging in each group was calculated. A significant association was found between poor labelling and the use of a Teflon catheter or butterfly needle for the injection of the stannous agent. In another 737 patients, in order to avoid the problems observed in the first group, a strict administration protocol was applied to analyse the frequency of poor tagging and its possible causes. Suboptimal image quality was present in 88 patients (11.9%). Quantitatively confirmed poor tagging was present in 36 of the 88 (40.9%, or 4.9% of the whole group); the remaining 52 patients showed borderline normal labelling (> 80% bound fraction). Drug interference was studied by comparing the medications used by the 36 patients showing poor binding with those used by 44 control patients. A significant relationship was found between the use of heparin or chemotherapy and the tagging. The influence of several clinical factors on the labelling was also investigated.(ABSTRACT TRUNCATED AT 250 WORDS)