The conductance catheter offers the only method for the determination of left ventricular volume on a continuous basis in patients. Two potential limitations of this technique include inhomogeneity of the electric field and current leakage into adjacent structures. The objectives of this study were to determine the extent of current leakage into adjacent structures and the sources of additional volume detected when the electric field is made more homogeneous. Both nonconductive and conductive cylinders over the range of left ventricular volumes seen clinically were constructed. The conductance catheter was radially centered in the models and a signal conditioner-processor was used to generate the electric field and calculate the corresponding volume. The conductive models were surrounded by saline to simulate surrounding structures. We demonstrate that electric current does not extend beyond the model wall, implying that adjacent structures such as the right ventricle are unlikely to affect the conductive volume measurements in vivo. Second,the increased homogeneity of the electric field detects more chamber but not more wall volume.