Atrial fibrillation is the most common clinical cardiac arrhythmia and is associated with cardiac dysfunction and stroke. It is believed to be triggered by ectopic electrical activity originating in the myocardial sleeves surrounding the pulmonary veins. Despite their importance, the processes within the pulmonary veins that lead to cardiac arrhythmias are unknown. The aim of this study was to investigate the architecture of the pulmonary veins. Male Sprague-Dawley rats (~450 g) were heparinized and humanely sacrificed in accordance with the UK Animals Scientific Procedures Act 1986. The hearts were perfused with phosphate buffered saline to flush out blood and then embedded in freezing medium and frozen in isopentane cooled in liquid N2. Cryosections of 20 μm thickness were cut in the coronal, sagittal and transverse planes from the whole hearts. From each heart, 16 sections (at intervals of 400 μm) were stained with Masson’s trichrome. At levels containing the pulmonary veins, sections were double immunolabelled for connexin43 (major gap junction protein in the working myocardium of the heart; Cx43) and caveolin3 (cardiac-specific isoform of the caveolin family) or alternatively Cx43 and HCN4 (major isoform responsible for the pacemaker current If in nodal tissues). The best view of the pulmonary veins was from sections cut in the transverse plane; it was possible to view the majority of the pulmonary veins in one section. Sections cut in the coronal plane were the best for showing the right superior pulmonary vein, as well as the sinoatrial and atrioventricular nodes. An example of a transverse section through the pulmonary veins immunolabelled for Cx43 and HCN4 is shown in Fig. 1. The marker of cardiac myocytes, caveolin3, was expressed throughout the atria (data not shown). It was also expressed in the myocardial sleeves of the pulmonary vein and its many branches (data not shown). The marker of working myocardium, Cx43, was also expressed throughout the atria (except for the sinoatrial and atrioventricular nodes) as well as the myocardial sleeves of the pulmonary vein and its branches (Fig. 1). In contrast, the nodal marker HCN4 was only expressed in the sinoatrial and atrioventricular nodes; it was not expressed in the working myocardium and it was not expressed at any point in the myocardial sleeves of the pulmonary vein and its branches (Fig. 1). HCN4 is, therefore, unlikely to be responsible for the ectopic activity of the pulmonary veins. In conclusion, it has been shown that the myocardial sleeves of the pulmonary veins can be identified by immunolabelling and the myocardial sleeves extend for many millimetres in the extensive pulmonary vein network of the rat. This is consistent with the work of Logantha et al. (1), who have shown cardiac action potentials can be recorded from intrapulmonary branches of vein.