Application of the current-clamp technique in rainbow trout atrial myocytes has yielded resting membrane potentials that are incompatible with normal atrial function. To investigate this paradox, we recorded the whole membrane current (I_m) and compared membrane potentials recorded in isolated cardiac myocytes and multicellular preparations. Atrial tissue and ventricular myocytes had stable resting potentials of -87±2 mV and -83.9±0.4 mV respectively. In contrast, 50 out of 59 atrial myocytes had unstable depolarized membrane potentials that were sensitive to the holding current. We hypothesized that this is at least partly due to a small slope conductance of I_m around the resting membrane potential in atrial myocytes. In accordance with this hypothesis the slope conductance of I_m was about 7 fold smaller in atrial than in ventricular myocytes. Interestingly, ACh increased I_m at -120 mV from 4.3 pA/pF to 27 pA/pF with an EC₅₀ of 45 nM in atrial myocytes. Moreover, 3 nM ACh increased the slope conductance of I_m four-fold, shifted its reversal potential from -78±3 to -84±3 mV and stabilized the resting membrane potential at -92±4 mV. ACh also shortened the action potential in both atrial myocytes and tissue and this effect was antagonized by atropine. When applied alone, atropine prolonged the action potential in atrial tissue but had no effects on membrane potential, action potential or membrane current in isolated atrial myocytes. This suggests that ACh-mediated activation of an inwardly rectifying K⁺-current can modulate the membrane potential in the trout atrial myocytes and stabilizes the resting membrane potential.