Conventional conductance volumetric systems require ex-vivo calibrations for blood conductivity and parallel conductance. It is often impractical to repeat blood sampling and hypertonic saline infusion for these calibrations. To overcome these limitations, we developed a useful, self-calibrating conductance volumetric system that does not require ex-vivo calibrations. On a conventional 6-electrode catheter, we added an extra electrode close to one of the recording electrodes to estimate blood conductivity. These two electrodes were placed close (0.5 mm) enough so that conductance between them reflected only blood conductivity regardless of cardiac volume. We estimated parallel conductance by the dual-frequency excitation (2 and 20 kHz) method. In 18 anesthetized rabbits, blood conductivity (sigma(est)) thus estimated agreed well with that (sigma(conv)) measured by the conventional ex-vivo blood sampling method (sigma(est) = 1.04sigma(conv)-0.25, R(2) = 0.98, SEE = 0.01 mS/cm, 1.2% error). Parallel conductance (G(p est)) estimated by dual-frequency excitation also agreed well with that (G(p conv)) estimated by the saline injection method (G(p est) = 0.95G(p conv)+4.25, R(2) = 0.87, SEE = 4.0 mS, 6.0% error). Estimated ventricular volume (V(est)) by our system agreed reasonably well with that (V(conv)) by the conventional method (V(est) = 0.93V(conv)+0.01, R(2) = 0.86, SEE = 0.22 ml, 14.7% error). The fact that this self-calibrating conductance volumetric system drastically simplifies volume measurement makes it an attractive tool for the assessment of cardiac function where significant changes in blood conductivity and parallel conductance are inevitable, such as in cardiac surgery.