The conductance catheter method has substantially enhanced the characterization of in vivo cardiovascular function in mice. Absolute volume determination requires assessment of parallel conductance (V(p)) offset because of conductivity of structures external to the blood pool. Although such a determination is achievable by hypertonic saline bolus injection, this method poses potential risks to mice because of volume loading and/or contractility changes. We tested another method based on differences between blood and muscle conductances at various catheter excitation frequencies (20 vs. 2 kHz) in 33 open-chest mice. The ratio of mean frequency-dependent signal difference to V(p) derived by hypertonic saline injection was consistent [0.095 +/- 0.01 (SD), n = 11], and both methods were strongly correlated (r(2) = 0.97, P < 0.0001). This correlation persisted when the ratio was prospectively applied to a separate group of animals (n = 12), with a combined regression relation of V(p(DF)) = 1.1 * V(p(Sal)) - 2.5 [where V(p(DF)) is V(p) derived by the dual-frequency method and V(p(Sal)) is V(p) derived by hypertonic saline bolus injection], r(2) = 0.95, standard error of the estimate = 1.1 microl, and mean difference = 0.6 +/- 1.4 microl. Varying V(p(Sal)) in a given animal resulted in parallel changes in V(p(DF)) (multiple regression r(2) = 0.92, P < 0.00001). The dominant source of V(p) in mice was found to be the left ventricular wall itself, since surrounding the heart in the chest with physiological saline or markedly varying right ventricular volumes had a minimal effect on the left ventricular volume signal. On the basis of V(p) and flow probe-derived cardiac output, end-diastolic volume and ejection fraction in normal mice were 28 +/- 3 microl and 81 +/- 6%, respectively, at a heart rate of 622 +/- 28 min(-1). Thus the dual-frequency method and independent flow signal can be used to provide absolute volumes in mice.