Background: The small conductance Ca2+-activated K+ current (ISK) is considered a potential target for preventing new onset atrial fibrillation (AF). The contribution of ISK to atrial repolarisation under physiological conditions is unclear, however, and substantial ISK activation may require [Ca2+]i exceeding typically measured global average diastolic-to-systolic levels. Aim: To test the hypothesis that ISK flows during the atrial action potential (AP), when local (sub-sarcolemmal, trigger) [Ca2+]i is expected to exceed global average levels, either at physiological or supra-physiological stimulation rates. Methods: Atrial myocytes were isolated from hearts excised from rabbits (anaesthetised with Na+-pentobarbital, 100 mg/kg I.V.), and from atrial tissues obtained from consenting patients who were in sinus rhythm and undergoing elective cardiac surgery. APs were recorded by whole-cell ruptured-patch clamp, at 35-37oC, and 1-3 Hz stimulation, before and after acute superfusion of an ISK blocker, ICA (1 or 10 µM). A +ve control, 4-AP (4-aminopyridine; K+ current blocker) was used initially since ISK, and thus the effect of ICA on APs, might be small. Results: 1) In rabbit atrial myocytes, stimulated at 1 Hz, 2 mM 4-AP significantly increased AP duration (APD): by 72% at APD30 (6.2±0.6 vs 3.6±0.3 ms; P<0.05), and by 31% at APD70 (21±4 vs 16±3 ms; P<0.05); with no significant effect on APD90 (n=6 cells, 2 rabbits). Drug responses were rapid (≤30 s), stable in all cells, and reversible in 5 of 6 cells studied. 2) By contrast, ICA, at 1 µM (~2 x IC50 for ISK) had no significant effect on APD30 (3.4±0.7 vs 3.7±0.6 ms), APD70 (27±9 vs 32±8 ms), or APD90 (82±14 vs 87±11 ms; n=10 cells, 4 rabbits), including relative to time-matched controls (TMC; 6 cells, 3 rabbits); nor on maximum diastolic potential (MDP) or AP maximum upstroke velocity (Vmax). 3) However, ICA at 10 µM (potentially non-selective for ISK) increased (non-reversibly) both APD70 and APD90 vs TMCs (by +2.9±1.2 ms vs -1.8±0.7 ms, and by +22±9 ms vs -3.9±2.1 ms, respectively; P<0.05 for each; 6-7 cells, 3 rabbits), also with no effect on MDP or Vmax. 4) In human atrial myocytes, also stimulated at 1 Hz, 1 µM ICA had no significant effect on APD30, APD70, or APD90 (n=9 cells, 4 patients), including relative to TMCs (4 cells, 2 patients), nor on MDP or Vmax. 5) At higher stimulation rates (2 or 3 Hz, for 80-110 s), intended to elevate [Ca2+]i and thus potentially enhance ISK, 1 µM ICA again had no significant effect on these AP parameters (n=7-10 cells, 3 patients), including vs TMCs (performed at 2 Hz; 6 cells, 3 patients). Conclusion: In ruptured-patch-clamped rabbit or human atrial isolated myocytes, ISK, assessed as any AP response to acute ICA at 1 µM, may not flow during AP stimulation at physiological rates, nor during short bursts of supra-physiological (up to 180 beats/min) stimulation. ISK activation (and thus its potential pharmacological inhibition during AF) may require changes to cellular electrophysiology or cell signalling systems to develop a sensitivity to ISK block.