Block of inactivating delayed rectifier K+ currents was studied in rat alveolar epithelial cells in primary culture using the whole-cell configuration of the gigohm-seal voltage-clamp technique. Charybdotoxin was the only blocker studied which did not alter K+ current kinetics; it produced a simple block (K1 approximately 1 nM) which appeared to be independent of voltage or channel state (open, closed or inactivated). Tetraethylammonium slowed inactivation of K+ currents, consistent with the notion that blocked channels cannot inactivate. Verapamil and methoxyverapamil produced time-, voltage- and concentration-dependent "inactivation" or block of open channels during depolarizing pulses, with negligible block of closed channels at negative holding potentials. Capsaicin, chlorpromazine, phencyclidine, quinidine and tetrahydroaminoacridine both increased the rate of inactivation and decreased the peak K+ current. These characteristics suggest that both open and closed channels can be blocked, but that open channels are blocked preferentially. Nifedipine, like most other blockers, increased the rate of K+ current decay, but, unlike other blockers, resulted in two distinct kinetic components of current decay under some conditions. Because nifedipine is uncharged, the voltage and time-dependence of its block cannot be ascribed to a traditional ionic blockade mechanism. Mechanisms of K+ channel block are compared with block of Ca++ channels by calcium "antagonists" and block of Na+ channels by local anesthetics. Interactions between gating kinetics and K+ channel blockade seem to be the rule rather than the exception.