In fifteen prospective patients with aortic stenosis undergoing transseptal cardiac catheterization, dye-dilution and count-based estimates of pulmonary blood volume (PBV) were performed. Three radionuclide methods were evaluated. Two were based on electrocardiogram (ECG)-gated imaging of the thorax, where pulmonary counts (PC) were corrected for frame-time, venous radioactivity, and either (1) the number of processed heart beats or (2) the total duration of acquisition. The third method involved ungated frame mode acquisitions, where PC were corrected for the duration of acquisition and the venous activity. PC (per channel element) were derived from manual assignments of the right lung. All methods correlated well with standard dye-dilution techniques (r greater than 0.82), though at greater volumes it was clear that count-based methods underestimated the dye-dilution values. In five acutely instrumented, anesthetized dogs, radionuclide (ungated formula) and dye-dilution estimates of PBV were made during multiple interventions (19 data points). The five control count volumes as well as the 14 separate intervention points correlated well (r greater than 0.89). It is concluded that PC from equilibrium blood pool images reflect PBV and that induced changes in PC can be utilized as a reflection of changes in PBV.