1. The dominant voltage-gated K+ currents in the somatic membrane of CA3 pyramidal cells from hippocampal slice cultures were characterized using the cell-attached configuration of the patch-clamp recording method. The kinetics, the voltage dependence of activation and inactivation, and the pharmacological properties of the current were determined from ensemble averages of large numbers of episodes from multichannel patches. 2. Steady-state analysis revealed that this current was half-inactivated at the resting membrane potential (Vr), and fully inactivated when patches were held 40 mV positive to Vr. Inactivation was removed when patches were hyperpolarized by 50 mV from Vr. Inactivation was well described by the Boltzmann equation with a slope factor of 12.6 mV. Removal of inactivation of the peak outward current could be described by a time-dependent monoexponential function with a time constant of the order of 100 ms. In contrast, the time course of inactivation was very slow: a +40 mV depolarization relative to Vr of several seconds was required for complete inactivation of the total outward current. 3. When steady-state inactivation was removed by hyperpolarization, the outward current activated with a threshold 10 mV positive to Vr and was half-activated at a potential 57 mV positive to Vr. The conductance can be described in terms of a single Boltzmann equation with a slope factor of 13.5 mV. Activation and inactivation properties of the somatic conductance produce a small window current between +10 and +20 mV relative to Vr. 4. The outward current activated in a voltage-dependent manner in less than 10 ms with 500 ms depolarizing steps. A kinetic analysis of its decay revealed at least three components, with the following time constants: a fast (17 ms), a slowly (approximately 150 ms), and a very slowly inactivating component (in the range of seconds). 5. External application of 4-aminopyridine (4-AP) induced a dose-dependent block of the peak outward current with an IC50 of 28 microM. The inhibitory effect of 4-AP saturated at a concentration of 200 microM which blocked 80% of the total current. The slowly and very slowly inactivating components of the current were not observed with 20 mM tetraethylammonium (TEA) in the pipette solution. A fast transient ensemble current (mean decay time constant, 24 ms) persisted in the presence of extracellular TEA in 29% of the patches. 6. In summary, at least two distinct voltage-gated K+ currents were present at the somatic level of hippocampal pyramidal cells. The dominant one, which we named IK(AT), is sensitive to micromolar concentrations of 4-AP and millimolar concentrations of TEA, and contributes three kinetic components to the total outward current. The second is TEA insensitive, and contributes only a fast transient component of outward current probably corresponding to the classic A-type K+ current. Intracellular recordings in CA3 pyramidal cells showed that IK(AT) plays an important role in regulating the duration of the action potential.