Gated blood pool radionuclide ventriculography provides a means for obtaining repeated studies of both cardiac ventricles with a single dose of radionuclide. Quantitative assessment of right ventricular (RV) function using this technique has been complicated by several technical problems. We describe a new method of RV blood pool analysis which attempts to solve these problems using well-established concepts for left ventricular (LV) blood pool analysis: (1) variable regions of interest; (2) computer edge detection with operator intervention; and (3) computer selected background. Results showed a strong linear correlation between gated first pass RV ejection fraction (RVEF) and the gated blood pool RVEF (n = 22; r = 0.93; blood pool RVEF = 0.03 + 0.89 X first pass RVEF; Sy.x = 0.04). There was also a strong linear correlation between LV and RV stroke counts in patients without valvular regurgitation, intracardiac shunts, or ventricular aneurysms (n = 19; r =0.86; RV counts = 72 + 0.94 X LV counts; Sy.x = 116). In terms of both of these validation standards this method proved superior to three published methods of RV blood pool analysis that used hand-drawn regions, and is suitable for analysis of rest, exercise, and intervention studies of RV function.