An arteriovenous fistula (AVF) is a vein graft which is created to permit access to the bloodstream allowing haemodialysis. AVFs are considered the best option in clinical practice; in spite of a 50% failure rate at 6 months (Field et al., 2008). Failure is primarily due to proliferation, leading to neointimal formation and invasion of the lumen (Lee and Roy-Chaudhury, 2009). Inflammation is known to play a key role in the development of vascular hyperproliferative remodelling, although the exact mechanisms and triggers are not fully known in AVF stenosis. The aims of this study were to; 1) evaluate the pathological changes undergone in human failed AVF vs. non-stenosed controls, 2) investigate the inflammatory processes undergone in human failed AVF vs. non-stenosed controls, 3) develop a rabbit model of AVF creation to study potential mechanisms that may contribute to stenosis. Histology of failed human AVF stained with H & E confirmed a significant increase in media: lumen ratio, from 2.2 ± 0.6 to 18.8 ± 6.9 in control and AVF vein sections respectively (p<0.005, Control vs. AVF; n=8). The majority of these cells where positive for α-smooth muscle actin (α-SMA). The percentage of cells undergoing proliferation, measured by expression of PCNA, increased from 3.5 ± 1.5% to 33.6 ± 4.9% in control and AVF vein sections respectively (p<0.005 AVF vs. Control; n=5, 8). The inflammatory profile of stenotic patients was measured by proteome array. Serum levels of pro-inflammatory molecules, including Il-6 and MCP-1 were significantly greater in Stenosed AVF patients vs. healthy control. Toluidine blue staining for mast cell infiltration highlighted a 5 fold increase in mast cells within AVF vein tissues (p<0.005 Control vs. AVF; n=9, 7), and infiltration of lymphocytes was evident from H & E staining. In our newly developed rabbit model, the animals were pre-medicated using hypnorm (IM, 0.3ml/Kg) 30 mins prior to surgery; and anesthetised with a mixture of isoflurane (1.5%), oxygen (1L min-1) and nitrous oxide (1L min-1). An AVF was then created between the femoral artery and vein, with Rimadyl (SC, 4mg/Kg) given as analgesia. tThe vein vein was monitored by ultrasound for 28 days post-creation of a femoral AVF to confirm patency. Animals were then pre-medicated (hypnorm, IM, 0.3ml/Kg) and euthanisedeuthanised (IV, euthatal, 1ml/Kg) and the vessels perfusion fixed in situ with paraformaldehyde. H & E staining of the vein showed the development of neointima (positive for α-SMA by immunohistochemistry), as well as neoadventitia within the venous segment. There was evidence of healing and integration at the anastomosis site, and a degree of neointima within the artery. Throughout the fistula, inflammatory cell infiltrates were present. These results highlight significant venous hypertrophic remodelling in human AVF-vein sections, largely attributed to vascular smooth muscle cell accumulation and inflammatory cell infiltrates. With successful development of the rabbit AVF model we are now in a good position to investigate the specific mechanisms leading to AVF stenosis.