Basic statistics textbooks invariably introduce the topic of prediction in the context of simple regression for the bivariate normal model.It seems more appropriate, however, to introduce the topic in conjunction with univariate normal inference.Prediction limits for the univariate case build on elementary theory that is familiar to the student.This approach provides an easier transition to prediction in the regression framework and exposes the student at an earlier stage to important applications and ideas connected with prediction.Adoption of the approach may give prediction the more prominent position in basic statistical education that it deserves.