The mechanisms responsible for atrial fibrillation (AF) are complex. At the tissue and cellular level, Ca2+ currents are down-regulated whereas repolarizing K+ currents are maintained, leading to action potential shortening. At the protein level however, Kv1.5 channels, the molecular expression of IKur, the main repolarizing current in human atria, is significantly decreased. We used a rat model of atrial remodeling induced by myocardial infarction (dilated atria) which reproduces the substrate of AF to investigate the mechanisms responsible for Kv1.5 channel internalization. Kv1.5 protein was decreased by 30% in dilated atria. However, the whole-cell current as measured by patch-clamp was maintained in dilated myocytes. The discrepancy between reduced protein level and maintained current density suggest that in dilated myocytes more Kv1.5 functional channels are inserted into the sarcolemma. We thus studied the internalization of Kv1.5 channels in control and diseased atria. Firstly, we investigated the endocytosis pathway for Kv1.5 channels in the atria. High resolution 3-Dimensionnal microscopy revealed that Kv1.5 channels are associated with clathrin vesicles in atrial myocytes but not with caveolin. Electron microscopy (EM) showed that vesicles are found both at the lateral sarcolemma and at the intercalated disc. Blockade of the clathrin pathway using hypertonic media or SiRNA induced an increase in IKur, an accumulation of Kv1.5 channels at the sarcolemma as shown by biotinylation assays and an increased fluorescence recovery after photobleaching (FRAP), supporting the idea that Kv1.5 channels are internalized via the clathrin pathway. Secondly, we studied the internalization activity in the dilated atria. In dilated atria, Western blot experiments showed a 35% reduction in clathrin levels. High resolution 3-Dimensionnal microscopy showed reduced association between Kv1.5 channels and clathrin vesicles in dilated myocytes. However, EM showed no significant difference in the internalization activity between sham and dilated atria. Therefore, the reduced clathrin protein synthesis observed in dilated atria is not likely to be responsible for the accumulation of Kv1.5 channels at the sarcolemma. Other mechanisms such as increased recycling and/or membrane stabilization must be investigated to understand how IKur is maintained in dilated atria and during chronic AF.