Inward currents carried by external Cs, Rb, NH4 and K through the IK1 channel were studied using a whole-cell voltage clamp technique. Cs, NH4, and Rb currents could be recorded negative to -40 mV following depolarizing prepulses (greater than or equal to 0 mV and 200-1000 msec in duration). The current activation displayed an instantaneous component followed by a monoexponential increase (tau a) to a peak amplitude. Subsequent inactivation was fit by a single exponential, tau ia. With hyperpolarization, tau a and tau ia decreased e-fold per 36 and 25 mV, respectively. In Ca-free external solutions (pipette [Mg] approximately 0.3 mM), inactivation was absent, consistent with the hypothesis that inactivation represents time- and voltage-dependent block of Cs, NH4, and Rb currents by external Ca. The inactivation and degree of steady-state block was greatest when Cs was the charge carrier, followed by NH4, and then Rb. K currents, however, did not inactivate in the presence of Ca. Na and Li did not carry any significant current within the resolution of our recordings. Comparison of peak inward current ratios (Ix/IK) as an index of permeability revealed a higher permeance of Cs (0.15), NH4 (0.30), and Rb (0.51) relative to K (1.0) than that obtained by comparing the steady-state current ratios (Cs:NH4:Rb:K approximately 0.01:0.06:0.21:1.0). At any given potential, tau a was smaller the more permeant the cation. In the absence of depolarizing prepulses, the amplitude of tau a was reduced. Divalent-free solutions did not significantly affect activation in the presence of 0.3 mM pipette [Mg]. When pipette [Mg] was buffered to approximately 50 microM, however, removal of external Ca and Mg lead to a four- to fivefold increase in Cs currents and loss of both time-dependent activation and inactivation (reversible upon repletion of external Ca). These results suggest that (i) permeability ratios for IK1 should account for differences in the degree to which monovalent currents are blocked by extracellular Ca and (ii) extracellular or intracellular divalent cations contribute to the slow phase of activation which may represent either (a) the actual rate of Mg or Ca extrusion from the channel into the cell, a process which may be enhanced by repulsive interaction with the incoming permeant monovalent cation or (b) an intrinsic gating process that is strongly modulated by the permeant monovalent ion and divalent cations.