Tonic inhibition is widely believed to be due solely to ‘spillover’ of GABA that escapes the synaptic cleft and activates extrasynaptic GABA_A receptors. However, an exclusively vesicular source is not consistent with the observation that tonic inhibition can still occur after blocking vesicular release. Here, we made patch clamp recordings from neurons in rat hippocampal cultures, and measured the tonic current that was blocked by bicuculline or gabazine. During perforated patch recordings, the tonic GABA current was decreased by the GAT1 antagonist SKF-89976a. Zero calcium solution did not change the amount of tonic current, despite a large reduction in vesicular GABA release. Perturbations that would be expected to alter the transmembrane sodium gradient influenced the tonic current. For example, in zero calcium Ringer, tetrodotoxin (which can decrease cytosolic [Na⁺]) reduced tonic current, whereas veratridine (which can increase cytosolic [Na⁺]) increased tonic current. Likewise, removal of extracellular sodium led to a large increase in tonic current. The increases in tonic current induced by veratridine and sodium removal were completely blocked by SKF89976a. When these experiments were repeated in hippocampal slices, similar results were obtained except that a GAT1- and GAT3-independent nonvesicular source(s) of GABA was found to contribute to the tonic current. We conclude that multiple sources can contribute to ambient GABA, including spillover and GAT1 reversal. The source of GABA release may be conceptually less important in determining the amount of tonic inhibition than the factors that control the equilibrium of GABA transporters such as the sodium gradient.