Regulation of large conductance, Ca²⁺– activated potassium (BK_Ca) channels by O₂-dependent enzymatic heme catalysis is primarily mediated by CO (Williams et al., 2004). However, the mechanism by which CO modulates BK_Ca channel activity remains elusive. Using HEK 293 cells stably and transiently expressing the α1 subunit (KCNMA1) of the BK_Ca channel, we have investigated the structural and molecular determinants of channel regulation by CO in excised, inside-out membrane patches. CO was introduced into the bathing solution by addition of 30 µM of a CO-releasing molecule or its stable, inactive analogue as a control. CO treatment resulted in a rapid and reversible activation of the BK_Ca whereas the control compound had no effect. Activation by CO was identical whether the β1 channel subunit (KCNMB1) was co-expressed with the α1-subunit or not. The effect of redox state of the channel was examined by treating patches with hypoxic (pO₂ ~ 40 mmHg) or hyperoxic (500 mmHg) solutions. Hypoxic conditions significantly reduced mean patch current from 0.40 ± 0.14 to 0.03 ± 0.01 pA (n=10, P<0.03). Hyperoxia had no significant effect on BK_Ca channel activity (P>0.3). Application of CO during either hypoxia or hyperoxia robustly increased mean patch current to 1.95 ± 0.80 pA (n =10, P < 0.05) and 1.18 ± 0.54 pA (n = 6, P < 0.05), respectively. It has been proposed (Jaggar et al., 2005) that CO regulates BKα channels via binding to reduced heme which associates with the α subunit at a consensus heme binding domain containing a conserved histidine (H611). Our data are not consistent with this hypothesis and we found that the redox state of exogenously added heme was unimportant to CO activation. Thus, in its oxidised form, heme (10 µM) was able to inhibit activity as predicted (from 0.35± 0.11 to 0.06 ± 0.02 (n = 14, P < 0.05), but concurrent CO treatment overcame this inhbition to increase mean patch current to 0.84 ± 0.36 pA (n = 14, P < 0.05). Moreover, heme itself (either reduced or oxidised) was not required for the CO effect since the channel activity of a H611R channel mutant (which does not bind heme, and was not inhibited by heme application – not shown) was still activated by CO. Finally, following the mutation of another histidine which has been implicated in CO regulation of the channels (H259R), singly and in combination with H611R, CO still activated both H259R (not shown) and H259R/H611R from 0.003 ± 0.002 to 0.178 ± 0.08 pA, n = 9, P < 0.05). These data show that, contrary to previous suggestions, CO binds to and activates BK_Ca channels by a mechanism which is independent of redox state, does not require heme and is independent of heme binding at conserved histidine residues.