1. Postnatal development of a large conductance Ca(2+)-activated K+ channel (BK channel) was investigated in neocortical infragranular pyramidal neurons with inside-out and outside-out patchclamp configurations. Neurons were acutely isolated from slices of 1- to 28-day-old rats (P1-P28) by using a vibrating glass probe after preincubation with low concentrations of enzymes. Patch membrane area was estimated by measuring membrane capacitance. The density, distribution, voltage dependence, Ca2+ sensitivity, kinetics, and pharmacological properties of BK channels were examined in neurons from animals of different ages. 2. In somata, the density of BK channels was 0.056 +/- 0.011/ micron 2 in P1 neurons and 0.312 +/- 0.008/micron 2 in P28 neurons. There was an abrupt increase between P5 and P7 at a rate of approximately 0.042/ micron 2/day. Before P5 and after P7, the density of BK channels also increased but at slower rates. 3. The density of BK channels in proximal apical dendrites underwent a similar developmental sequence. There was a relatively large increase between P5 and P7 with a rate of approximately 0.021/ micron 2/day, and after P7, channel density increased more slowly (approximately 0.002/microns 2/day). In P1 neurons, channel density in apical dendrites was 0.039 +/- 0.008/micron 2, which was close to that in somata, whereas in P28 neurons, channel density (0.134 +/- 0.008/micron 2) was less than one-half of that in somata. 4. The distribution of BK channels was different in immature and mature neurons. In somata of P1 neurons, BK channels were distributed singly without evidence of clustering, whereas in P28 neurons BK channels were clustered in groups of approximately 4. 5. BK channels in both P1 and P14 neurons showed a steep increase in the probability of opening (Po) as intracellular Ca2+ concentration was raised from 50 to 100 nM, especially at positive membrane potentials. The Ca2+ dependence, as measured by the [Ca2+]i that provided half-maximal Po at a variety of membrane potentials, was not different in patches from P1 and P14 neurons. On the other hand, the voltage dependence of BK channels shifted during ontogeny such that Po was larger at negative potentials in P14 than in P1 neurons. 6. The voltage dependence of P1 BK channels was bimodally distributed with 57% of channels exhibiting an "immature" pattern consisting of a more positive V1/2 and a smaller change in voltage required to produce an e-fold increase in Po. Immature type P1 BK channels showed a longer mean closed time at negative membrane potentials than either P14 or "mature" P1 BK channels. 7. No postnatal developmental changes in pharmacological properties of BK channels were observed. In both mature and immature neurons, BK channels were partially inhibited by 30 or 100 nM charybdotoxin (ChTX) and fully blocked by 1 microM ChTX. The IC50 for ChTX was 100 nM, indicating that BK channels in neocortical pyramidal neurons are much less sensitive to ChTX than those in muscle cells and sympathetic ganglion neurons. BK channels were also inhibited by 0.5 mM tetraethylammonium chloride (TEA) and 50 microM trifluoperazine. 8. These data indicate that functional somatic and dendritic BK channels are inserted into neuronal membranes during neocortical development, with an especially rapid increment in density occurring around P5-P7. These changes, which occur at a time when other voltage-gated ion channels are known to be increasing in density, contribute to the development of neocortical excitability.