The number of patients with end stage renal disease is still increasing. Most of them are treated by hemodialysis. This method requires permanent vascular access (VA). The best type is arteriovenous fistula but the number of used subtypes is very small. Therefore the lenght of survival of each single access becomes crucial. A review of the current literature on the planning of VA, access surveillance methods, neointimal hyperplasia (NIH) and treatment of the most common complications was performed. Multidisciplinary VA planning and increased use of preoperative imaging are essential to obtain the best long-term results. The research of shunt surveillance is still at the beginning. We are able to detect venous stenosis early but the precise indications for intervention remain unclear. NIH plays important role in the genesis of VA complications. The only method which reduces NIH and can soon improve survival of shunts in clinical practice is suturing of anastomoses with nonpenetrating vascular clips. Management of complications as performed today is successful and with good outcome. Steal syndrome is now more frequent complication, especially by patients with diabetes and atherosclerosis. New methods of solving this complication without access ligation are now developed. Research focuses on optimalization of all procedures of VA creation to improve the lenght of survival of each access. Each improvement of secondary access patency is important as a means of prolonging patient survival, reducing morbidity and also cost of hemodialysis programmes.