In the desert of eastern Washington State, about 350 km from Seattle, rests more than 200000 m/sup 3/ Of radioactive and chemically hazardous waste. A legacy of over 50 years of nuclear weapons production, the waste is a mix of liquids, solids, sludges, and gases that sits in 177 aging steel tanks. So far more than a third of these tanks, which are buried underground, have leaked or are suspected to have done so, releasing an estimated 3800 m/sup 3/ of hazardous radioactive materials into the soil. Some of this leakage has seeped more than 60 meters down into groundwater about 10 km from the Columbia River, the most important waterway in the northwestern United States. For more than 10 years, the US Department of Energy (DOE) has tried to devise a plan to treat this waste, first to stabilize it and then to dispose of it. To date, more than U S $5 billion has been spent on this challenge, with little to show for it. Construction of facilities to treat the waste has not yet begun, and the history of this radioactive cleanup has been punctuated by failed designs, false starts, a variety of technical and financial problems, and changing policies. By DOE's own calculations, to complete this project, it will have to spend roughly $1 billion a year at Hanford for the next 40 years. The reasons for this state of affairs are complex. They involve technical and managerial shortcomings in the face of unique challenges that are unlike anything else the world has ever seen. The tank waste itself is extremely difficult to retrieve, and can be safely manipulated or sampled only by workers who are shielded and trained.