Science-based stockpile stewardship is the current program undertaken by the United States Department of Energy to test the reliability of its nuclear weapons stockpile. The US is currently a signatory of the Comprehensive Test Ban Treaty or CTBT, although it has not ratified it. Nuclear-capable signatories to this treaty are not allowed to test nuclear weapons systems at any yield. In order to test the reliability of its current nuclear weapons systems, the United States uses computer simulation and materials testing to certify the integrity of its arsenal.

Nuclear weapons are inherently unstable machines. Their main ingredients, plutonium and, in thermonuclear weapons, tritium, decay over time due to their radioactivity. This decay can result in changes to the metallurgical properties of the material, and can also introduce decay products that poison nuclear reactions and reduce the designed yield of the device. The high explosive used in primary implosion can also degrade over time, and the weapons themselves may be subjected to many intentional and unintentional stresses which may damage one or more components of the system. Science-based stockpile stewardship attempts to quantify the impact of these factors on weapon performance.

SBSS is one of the primary reasons for the existence of the US nuclear weapons laboratories of Los Alamos National Laboratory, Lawrence Livermore National Laboratory, and Sandia National Laboratory. Every year, the directors of those laboratories are legally required to certify to the President and Congress that the nuclear arsenal of the United States is in proper working order. The scientific work of SBSS is to ensure that it is.

All of the nuclear capable states must have some level of SBSS integrated into their weapons programs. In the United Kingdom, this role is filled by the Atomic Weapons Establishment (AWE), and in France, by the Direction des Applications Militaires (DAM) of the Commissariat à l'Energie Atomique. The United States has worked with the UK on matters related to nuclear weapons science since the Manhattan Project, and formalized this relationship in a 1958 agreement on mutual defense. The United States conducts a small amount of research in collaboration with the Russian nuclear agency, but has no official collaboration with any other nuclear powers.

Several questions exist as to the validity of the SBSS approach, including how well the degradation of materials over time is understood, and whether remanufacturing of questionable weapons systems within the framework of existing arms control agreements would be a safer and more cost-effective approach. Essentially, SBSS is a trade-off. Not testing nuclear weapons encourages proliferant nuclear states not to test or develop their own weapons, but it doesn't allow nations which practice SBSS to develop new weapons or upgrade their old ones. Without SBSS nuclear states would either have to conduct nuclear tests, or maintain the industrial capability to build new ones (or unilaterally disarm).