1. The time-and voltage-dependent properties of a slowly inactivating and depolarization-activated potassium current and the functional consequences of its activation was investigated with current and single-electrode voltage-clamp techniques applied to guinea pig dorsal lateral geniculate neurons maintained as a slice in vitro. 2. In current clamp, application of a step depolarization to near firing threshold resulted in a slowly rising membrane potential that took up to 10 s to reach steady state and firing threshold. In voltage clamp, step depolarization of the membrane potential to values positive to approximately -65 mV resulted in the rapid activation followed by slow inactivation of an outward current. In both cases the sudden depolarization was associated with a large increase in membrane conductance, which gradually lessened in parallel with the slow depolarization in current clamp or with the decrease in outward current in voltage clamp. 3. The time course of inactivation of the outward current, which we refer to as I(As), was well fitted by a two-exponential function with time constants of 96 and 2,255 ms, suggesting the presence of a fast and slow phase of inactivation. The activation threshold for I(As) was about -65 to -60 mV, whereas inactivation was incomplete even at -50 mV, suggesting the presence of a substantial "window" current. The time course of removal of inactivation of I(As) at -85 to - 100 mV was well fitted by a single exponential function with time constant of 91 ms. 4. I(As) appears to be mediated by K+. Increasing [K+]o from 2.5 to 10 mM resulted in a reduction in amplitude of I(As), whereas changing from 10 to 2.5 mM [K+]0 enhanced this current. Intracellular injection of Cs+ resulted in an abolition of I(As), whereas extracellular application of Ba2+ resulted in a large decrease in the apparent input conductance but relatively little reduction of I(As). 5. Both phases of inactivation of the transient outward current were completely blocked by low doses (100-mu-M) of 4-aminopyridine (4-AP), but not by extracellular application of Cs+, tetraethylammonium (TEA), tetrodotoxin (TTX), or after block of transmembrane Ca2+ currents. Local application of 4-AP to neurons depolarized to near firing threshold resulted in depolarization associated with a decrease in apparent input conductance, thereby confirming the presence of a window current. 6. The fast activation and slow inactivation characteristics of I(As) resulted in substantial rectification of the membrane such that the cell's response to short-duration (< 100 ms) depolarizing current pulses was much smaller than that to hyperpolarizing current pulses of the same magnitude. Block of I(AS) with 4-AP, or tonic depolarization of the membrane potential positive to -55 mV, reduced this bias against depolarizing inputs. 7. Block of I(As) with 4-AP resulted in a substantial increase in peak amplitude and duration of low-threshold Ca2+ spikes, indicating that this transient potassium current participates in controlling intrinsic burst firing. 8. These properties of I(As) bias thalamocortical relay neurons against transient depolarizations from a relatively hyperpolarized membrane potential (-65 mV), although leaving the cells responsiveness to transient hyperpolarizations largely intact. In contrast, after tonic depolarization of the membrane potential, this bias is largely lost. These functional consequences of I(As) may contribute substantially to the ability of thalamocortical relay neurons to rhythmically oscillate in response to inhibitory postsynaptic potentials although being less sensitive to excitatory postsynaptic potentials during periods of drowsiness and slow-wave sleep. In contrast, the tonic depolarization of thalamocortical relay neurons, which is known to occur during arousal and attentiveness, will reduce the influence of I(As) through inactivation, and the bias against excitatory postsynaptic potentials will lessen, thereby increasing the accurate relay of sensory information.