The measure of membrane capacitance (Cm) in cardiac myocytes is of primary importance as an index of their size in physiological and pathological conditions, and for the understanding of their excitability. Although a plethora of very accurate methods has been developed to access Cm value in single cells, cardiac electrophysiologists still use, in the majority of laboratories, classical direct current techniques as they have been established in the early days of cardiac cellular electrophysiology. These techniques are based on the assumption that cardiac membrane resistance (Rm) is constant, or changes negligibly, in a narrow potential range around resting potential. Using patch-clamp whole-cell recordings, both in current-clamp and voltage-clamp conditions, and numerical simulations, we document here the voltage-dependency of Rm, up to −45% of its resting value for 10-mV hyperpolarization, in resting rat ventricular myocytes. We show how this dependency makes classical protocols to misestimate Cm in a voltage-dependent manner (up to 20% errors), which can dramatically affect Cm-based calculations on cell size and on intracellular ion dynamics. We develop a simple mechanistic model to fit experimental data and obtain voltage-independent estimates of Cm, and we show that accurate estimates can also be extrapolated from the classical approach.