An extension of ordinary least squares (OLS) estimation of a linear regression in the presence of endogeneity, that is, when the error term is correlated with explanatory variables. In the first stage, the endogenous explanatory variables are regressed on appropriately chosen instrumental variables, using OLS. In the second stage, the original regression is estimated, using OLS, with the endogenous explanatory variables replaced by their fitted values from the first stage. Under certain conditions the 2SLS (or IV) estimator is consistent and efficient. See also Hausman test.