ATP-sensitive potassium (K_ATP) channel closure in response to glucose metabolism links changes in blood glucose concentration to insulin secretion from pancreatic β-cells. K_ATP channels are inhibited by ATP (which binds to the Kir6.2 subunit) and activated by Mg-nucleotides such as MgADP and MgATP which interact with the SUR1 subunit. In excised membrane patches, the channel is half-maximally blocked (IC₅₀) by 10-30μM ATP and fully blocked by 1 mM ATP. However, channel activity is observed in on-cell patches on β-cells when [ATP]_i is predicted to be 1-5 mM. This suggests the apparent ATP sensitivity in the excised patch differs from that in intact β-cells, due to loss of cytoskeletal elements or cytosolic modulators such as MgADP or PIP₂, which enhance K_ATP channel activity and decrease its ATP sensitivity. To address this issue, we established an open-cell-attached patch-clamp configuration (open-cell) that enables K_ATP channel ATP sensitivity to be measured in cell-attached patches on permeabilized cells. B-cells were isolated from pancreata of humanely killed NMRI mice and cultured for 1-3 days. Single-channel currents were recorded at -60 mV using the open-cell configuration. Cells were permeabilized with 5 μg/ml S. aureus α-toxin or 20-30 HU/ml streptolysin O (SLO). The pipette solution contained (mM): 140 KCl, 10 HEPES (pH 7.2 with KOH), 1.1 MgCl₂, 2.6 CaCl₂. The bath solution contained (mM): 30 KCl, 110 K-aspartate, 0.084 CaCl₂, 2 MgSO₄, 10 HEPES (pH 7.2 with KOH), 0.5 EGTA and various [ATP]. The Mg²⁺-free bath solution contained (mM): 30 KCl, 110 K-aspartate, 2.6 CaCl₂, 10 HEPES (pH 7.2 with KOH), 0.5 EGTA, 5 EDTA. Data are given as mean±SEM; n is the number of patches. The ATP sensitivity of the K_ATP channel in the open-cell configuration was markedly lower than in the excised patch. The IC₅₀ was 101±28 μM (n=6) in open-cell vs 23±2 μM (n=5) in excised patch (p<0.05), when permeabilized using α-toxin (cut-off <3 kDa). There was no difference in the open-cell ATP sensitivty when SLO (cut-off 150 kDa) was used for permeabilization (IC₅₀ = 109±38 μM (n=7)), suggesting that it is not due to readily diffusible agents of low molecular weight. However, the difference in ATP-sensitivity was abolished in Mg²⁺-free solution (IC₅₀ = 14±4 μM (n=6) in open-cell and 10±1 μM (n=5) in excised patch) or when the poorly hydrolysable ATP analogue, ATPγS, was used (IC₅₀ = 7±1 μM (n=6) in open-cell, 5±1 μM (n=5) in excised patch). These data suggest a Mg²⁺-dependent process underlies the reduction in apparent ATP sensitivity seen in open-cell patches. Possible mechanisms include Mg-nucleotide activation via SUR1, and ATP-dependent generation of PIP₃/PIP₂. A contribution of the latter is suggested, as LY294002 (100μM), which blocks PIP₂ synthesis, reduced the ATP sensitivity in both open-cell and excised patch (IC₅₀ = 21±3 μM (n=8) in open-cell, 9±1 μM (n=5) in excised patch).