Cerebrospinal fluid (CSF) shunts were invented almost 50 years ago. While their introduction revolutionized the treatment of hydrocephalus, their complications have become legendary, and the focus of much investigation and development of new devices. New devices have been based upon improved understanding of the pathophysiology of hydrocephalus or shunt complications. Despite the rational, or frequently "more physiological," functioning of these devices, all too often unexpected complications have ensued, and the initial enthusiasm for the devices has waned. Assessing the efficacy of the devices has been difficult, owing to the lack of properly conducted studies. Nevertheless, the overall impact of shunt design improvements has seemed very limited. A recent randomized trial of CSF shunt design, examining the failure rates of two new and widely used valves (the Cordis Orbis Sigma and the Medtronic PS Medical Delta valves) failed to find any advantage of these over standard valve designs, many of which have been used almost since the inception of CSF shunts. A search for risk factors for failure, in a post hoc analysis of the data, indicated only that the etiology of the hydrocephalus and the position and local environment of the ventricular catheter tip were probably important. Remarkably, the rate of change in the size of the ventricles and the final ventricular size were not different despite the substantial differences in flow characteristics of the two new valves. Shunt failure rates of less than 5% at 1 year, with infection rates of less than 1%, seem like reasonable goals for the next decade in the new millenium. This can be achieved through basic research into the pathophysiology of shunt failure with improved mathematical models, and perhaps animal models of shunt failure. Efficacy of new devices or treatments must be scrutinized scientifically so as not to waste valuable resources and time on unproven treatments. Uncontrolled series and testimonial assertions about new treatments or devices, especially from proponents with a vested interest, should be regarded with great skepticism. Nevertheless, our best efforts are likely to result in a major advance in the management of pediatric hydrocephalus, which now seems tantalizingly close.