The fitted values are the predicted values of the $y_is according to the model. $\hat y_i = \hat \beta_0 + \hat \beta_1 X_{i, 1} + \dots + \hat \beta_p X_{i, p}$ where $\hat \beta_j$ is the [[least squares estimator]] for $j= 0, \dots, p$. ```R predict(lm_model, newdata=x) ``` See also [[prediction interval]].