To test the hypothesis that general anesthetics block nicotinic acetylcholine receptor channels by binding within the pore of the channel, we looked for competitive interactions between ether and QX-222 at the single channel current level. Experiments were performed on outside-out patches excised from BC3H-1 cells. QX-222 causes channels to flicker as it repeatedly binds within the pore of the channel and blocks the flow of current through the channel. Ether reduces the apparent unitary conductance of the channel. This effect of ether may be due to frequent, short-lived, unresolved, blockages of the channel. When both ether and QX-222 are applied, the effects of both drugs are seen on single channels. However, the duration of QX-222 blocking events are longer when ether is present; the duration of block is 0.89 +/- 0.06 ms with 30 microM QX-222 alone and 2.23 +/- 0.37 ms with 30 microM QX-222 + 20 mM ether (n = 5 +/- S.D.; -100 mV). Similar results are obtained when butanol is used in place of ether. We conclude that ether and QX-222 do not compete for a common binding site. Conversely, ether decreases the dissociation rate of QX-222. The simplest interpretation of these data is that the binding sites for ether and the aromatic moiety of QX-222 are distinct but close to each other; when ether is bound to its site, the binding of QX-222 is stabilized. We cannot, however, discount the possibility that ether stabilizes QX-222 by binding to a remote site and allosterically modifying the pore of the channel.