Decision analysis is an intellectual technology for organizing choices involving multiple competing costs and benefits, choices involving them, an the uncertainty in how choices lead to different outcomes. Decision analysis is a modern form of what Aristotle calls practical reason, meaning methods used to determine how best to achieve valued goals given your resources.

Decision analysis combines methods from mathematics and behavioral science. The mathematical framework provides a toolkit of ideas for structuring complicated decisions into manageable pieces, formalizing uncertainty, and making sense of complex tradeoffs. There is no mathematical technique for averaging political differences and incommensurable values, but there are ways to make those more explicit, transparent and open to debate.

The behavioral side of decision analysis takes account of the challenges in judgment and choice everyone faces when dealing with complexity and uncertainty, including "experts" and engineers. On the one hand, there are predictable biases and errors people make which can be avoided, especially dealing with the concepts of chance and risk. At the same time, many public decisions are so complex they will always have to be simplified and streamlined for any reasonable decision-making process. That applies equally to city councils, land use commissions and scientific academies. In such situations decision analysis provides useful means for organizing measurement issues, modeling, the expression of uncertainty, and the comparison of alternatives

Decision analysis differs from risk analysis by considering the choices one has available to make a difference in the world, not just what makes the world a dangerous place.

A very useful part of decision analysis is called multiple values decision analysis, or in technical jargon, multiattribute utility theory. It deals with situations in which there are multiple competing outcomes in a decision, such as human health impacts, ecological changes, dollar losses, jobs, aesthetic changes, and so on, as is typical in many environmental problems. The theory offers an alternative to economic theories which monetize all values. Instead incommensurable values are kept distinct, with comparisons tracked directly through choices and tradeoffs. The theory also has natural techniques for constructing novel measurement criteria and organizing public involvement processes.