We examined the effect of the Ca²⁺ concentration in the endolymph ([Ca]_e) or in the endolymphatic surface cells ([Ca]_i) on the endocochlear potential (EP) by using an endolymphatic or perilymphatic perfusion technique, respectively. (i) A large increase in [Ca]_e up to ∼10⁻³ M with a fall in the EP was induced by transient asphyxia (∼2 min) or by the intravenous administration of furosemide (60 mg/kg), and a significant correlation was obtained between the EP and p[Ca]_e (= −log [Ca]_e, r = 0.998). (ii) Perfusion of the endolymph with 10 mM EGTA for 5 min neither produced any significant change in the EP nor altered the asphyxia-induced change in EP (ΔEP_asp), suggesting that neither [Ca]_e nor the Ca²⁺ concentration gradient across the stria vascularis contributed directly to the generation of the EP in the condition of low [Ca]_e. In contrast, endolymphatic perfusion with high Ca²⁺ (more than 10 mM) produced a decrease in EP and a significant correlation was obtained between the EP and the Ca²⁺ concentration of perfusion solution (r = 0.982), suggesting that Ca²⁺ permeability may exist across the stria vascularis. (iii) The administration of a Ca²⁺ chelator, EGTA-acetoxymethyl ester (AM, 0.3 mM), to the endolymph, which produced a gradual increase in EP, suppressed significantly, by 60–80%, ΔEP_asp or furosemide-induced changes in EP. In contrast, perilymphatic administration of 0.5 mM EGTA-AM caused no significant suppression of the ΔEP_asp. These findings suggest that [Ca]_i plays an important role in generating/maintaining a large positive EP.