Issues involved in standard setting along with methods for standard setting are reviewed, with specific reference to their relevance for criterion referenced testing. Definitions are given of continuum and state models, and traditional and normative standard setting procedures. Since continuum models are considered more appropriate for criterion referenced testing purposes, they are examined in greater depth. The continuum models are subdivided into three categories: judgmental methods; empirical methods; combinations of judgment and empirical methods. For the purpose of viewing a test as an entity and not in relation to other variables, the Angoff and Nedelsky judgmental methods are considered useful; however, when empirical data is available, the empirical methods delineated in Berk's method of the Contrasting Groups method are recommended as being appropriate. Discussed in the final section of the paper are procedures for setting standards to accomplish three primary uses of criterion referenced testing: classroom testing; basic skills testing for annual promotion and high school graduation; professional licensing and certification testing. However, further research is recommended before the latter procedures are implemented. (Author/AEF)