We describe a model for the mechanism by which Ca channel inactivation might depend on calcium entry. Ca is assumed to bind to a site at the internal membrane surface to cause inactivation of Ca channels. We assume that Ca that enters through the membrane accumulates in a submembrane compartment and also make simplifying assumptions about Ca buffering and removal. Our model predicts the results of single-and double-pulse voltage-clamp experiments well. The predicted turn-off of Ca current is non-exponential. The model also predicts that procedures that slow inactivation will increase peak Ca current and suggests that both two-phase turn-off of currents and failure of normalized current to recover to 1.0 in two-pulse experiments may be explained without assuming a voltage-dependent component of inactivation or two populations of Ca channels.