The document was prepared in support of an advisory process meeting to produce a technical guidance document to assist science practitioners responsible for developing the science elements of the Precautionary Approach (PA) framework. It reviews the use of reference points in the assessment and management of semelparous species and anadromous salmonids including iteroparous salmon. Semelparous species and anadromous salmonids are treated collectively and separately from other aquatic species because they share a number of life history and population dynamic characteristics which are distinct from those of other aquatic organisms. In British Columbia and Yukon, the Wild Salmon Policy (WSP) guides the implementation of the precautionary approach in the management of fisheries on Pacific salmon. In the WSP, biological benchmarks are developed for four main classes of indicators: trends in abundance, abundance, fishing mortality, and spawning ground distribution as data permits. These indicators are further integrated into a single category of biological status. There are no management actions which are associated directly with a given status or benchmark. Rather, the biological benchmarks aid in the development of fishery reference points along with socio-economic factors and issues related to risk tolerance. Reference points for Atlantic salmon have been used to advise fisheries management since the 1970s. The use of a conservation objective defined as a limit reference point and the fixed escapement strategy has been adopted in Canada, by national governments in Europe and by international organisations. Candidate fishery reference points and WSP benchmarks are similar to the general list of reference points proposed in a number of publications for other species. Stock and recruitment models have a long and established history in Pacific and Atlantic salmon stock assessment and provision of science advice for fisheries management. Empirical methods consisting of life history models that use life history process parameters borrowed from a large range of studies on the species of interest are considered. In data limited situations for unstudied populations but for which information exists from other populations, reference points are frequently transported based on values from studied populations which are standardized using an exchangeable and transportable metric.