Many genes change their transcription in atrial fibrillation (AF) patients, which indicates the functional relationship of these genes with AF. Thus, identifying the over-represented motifs in the upstream region of these genes would shed light on the mechanism of this disease. We analyzed a set of microarray experiments, selected from genes whose expression was changed, and extracted their regulatory sequences from genome databases. In addition, we used a motif discovery algorithm to obtain frequent DNA motifs in these sequences. Representative motifs were selected and matched to known transcription factor binding sites by clustering analysis, and 3 putative motifs related to AF were finally identified.