The aortic valve is a beautifully designed and balanced structure with precise opening and closing mechanisms. Its design probably cannot be improved upon by man-made mechanical devices, no matter how ingenious. It is logical to use human valves as replacements because of their superior structural configuration. Their chief difficulties are in the troubles of the learning curve of the surgeon to achieve a competent mechanism and the hazard of infection. The pulmonary autograft or translocation of the pulmonary valve to the aortic position offers many advantages, and the long-term results are good. Conservative repair and reconstruction of the aortic valve are increasing in popularity but have a relatively limited application. The mitral valve has a less precise and more adjustable type of mechanism and generous leaflet coaptation. These features make it more amenable to repair, as demonstrated in the clinical series of the technique. Repair is clearly preferable to replacement at present. The tricuspid valve is eminently suitable for reconstruction, but late results are not very good. The disappointing outcome, however, reflects more the severity of underlying cardiac pathology than the effectiveness of the repair.