Atrial fibrillation is the most common arrhythmia at alcohol consumption. Its pathogenesis is complex, at least partly related to changes of cardiac inward rectifier potassium currents including I-K1. Both ethanol and acetaldehyde have been demonstrated to considerably modify I-K1 in rat ventricular myocytes. However, analogical data on the atrial I-K1 are lacking. The present study aimed to analyse I-K1 changes induced by ethanol and acetyldehyde in atrial myocytes. The experiments were performed by the whole cell patch-clamp technique at 23 +/- 1 degrees C on enzymatically isolated rat and guinea-pig atrial myocytes as well as on expressed human Kir2.3 channels. Ethanol (8 - 80 mM) caused a dual effect on the atrial I-k1 showing the steady-state activation in some cells but inhibition in others in agreement with the ventricular data; on average, the activation was observed (at 20 mM by 4.3 and 4.5% in rat and guinea-pig atrial myocytes, respectively). The effect slightly increased with depolarization above -60 mV. In contrast, the current through human Kir2.3 channels (prevailing atrial I-K1 subunit) was inhibited in all measured cells. Unlike ethanol, acetaldehyde (3 mu M) markedly inhibited the rat atrial I-K1 (by 15.1%) in a voltage-independent manner, comparably to the rat ventricular I-K1. The concurrent application of ethanol (20 mM) and acetaldehyde (3 mu M) resulted in the steady-state I-K1 activation by 2.1% on average. We conclude that ethanol and even more acetaldehyde affected I-K1 at clinically relevant concentrations if applied separately. Their combined effect did not significantly differ from the effect of ethanol alone.