Attempts to determine the appropriate value for a currency pair entail the use of regressions. In this chapter, three types of regression models are introduced. One note of caution regarding the use and interpretation of long‐term regressions takes into account the unstable relationships between the dependent and independent variables. In conclusion, the long‐term, fair‐value regression model for EUR/USD provides intuitively appealing explanatory variables: the 10‐year yield spread, the 10‐year average yield, and the CPI inflation ratio. This chapter has explored fair‐value regression estimates for currencies based on valuations assessed over varied time periods. At the highest level, fair‐value regression analysis allows an analyst to quantify from a high‐level, long‐term view when a currency pair is over‐ or undervalued. Regression analysis based on 52 weeks of weekly closing data can be used to ‘zoom in’ from the long‐term analysis in order to select more frequent investment and trading opportunities.