Activation of large conductance Ca²⁺-activated K⁺ channels is controlled by both cytoplasmic Ca²⁺ and membrane potential. To study the mechanism of voltage-dependent gating, we examined mSlo Ca²⁺-activated K⁺ currents in excised macropatches from Xenopus oocytes in the virtual absence of Ca²⁺ (<1 nM). In response to a voltage step, I_K activates with an exponential time course, following a brief delay. The delay suggests that rapid transitions precede channel opening. The later exponential time course suggests that activation also involves a slower rate-limiting step. However, the time constant of I_K relaxation [τ(I_K)] exhibits a complex voltage dependence that is inconsistent with models that contain a single rate limiting step. τ(I_K) increases weakly with voltage from −500 to −20 mV, with an equivalent charge (z) of only 0.14 e, and displays a stronger voltage dependence from +30 to +140 mV (z = 0.49 e), which then decreases from +180 to +240 mV (z = −0.29 e). Similarly, the steady state G_K–V relationship exhibits a maximum voltage dependence (z = 2 e) from 0 to +100 mV, and is weakly voltage dependent (z ≅ 0.4 e) at more negative voltages, where P_o = 10⁻⁵–10⁻⁶. These results can be understood in terms of a gating scheme where a central transition between a closed and an open conformation is allosterically regulated by the state of four independent and identical voltage sensors. In the absence of Ca²⁺, this allosteric mechanism results in a gating scheme with five closed (C) and five open (O) states, where the majority of the channel's voltage dependence results from rapid C–C and O–O transitions, whereas the C–O transitions are rate limiting and weakly voltage dependent. These conclusions not only provide a framework for interpreting studies of large conductance Ca²⁺-activated K⁺ channel voltage gating, but also have important implications for understanding the mechanism of Ca²⁺ sensitivity.