Radionuclide ventriculography is a useful investigation in the evaluation of cardiac function. Generally, in vivo technetium 99m-labelled red blood cells (RBC) yield good quality images in ventriculography. However, it is widely believed that some drugs have an adverse effect on RBC labelling. Zanelli et al. (1987) developed a radiopharmaceutical (technetium 99m bis-diethylphosphinoethanebis-t-butylisocyanide, 99mTc-DEPIC) to obtain better results in patients using such drugs. We undertook a prospective study of 6 patients with cardiovascular and/or pulmonary disease using several kinds of drugs to evaluate imaging of the cardiac blood pool with 99mTc-DEPIC and in vivo labelled 99mTc-RBC. After injection, blood samples were taken, and gated equilibrium blood pool studies were performed. The radiochemical purity of the injected 99mTc-DEPIC varied from 76.4 to 93.6% (mean 86.4%, SD 5.7%). The protein (pre-albumin) binding was 100%. Biological half-life in blood varied from 3.3 to 4.7 h (mean 4.1 h, SD 0.5 h). For 99mTc-RBC no significant blood disappearance was seen for 8 h. The percentage of RBC-bound 99mTc varied from 96.9% to 98.3% (mean 97.0%, SD 0.5%) and was stable for at least 8 h. The heart-to-lung, heart-to-spleen, and heart-to-liver ratios were higher for 99mTc-RBC than for 99mTc-DEPIC. Furthermore, 99mTc-DEPIC showed a significant decline of the ejection fraction with time. Visually, the images with 99mTc-RBC were superior to those with 99mTc-DEPIC, especially a few hours after injection. According to our findings, in vivo labelling of 99mTc-RBC is still the method of choice for routine radionuclide ventriculography.(ABSTRACT TRUNCATED AT 250 WORDS)