I have my dear copy of Abramowitz and Stegun on the bookshelf next to my desk and I refer to it nearly weekly. E.g., yesterday I had a postdoc asking me about a strange numerical scheme for computing derivatives using five points that he found, and looking at the appropriate page on the handbook we could easily find its name, history, and main properties; and thanks to that, understand whether a different scheme would be more appropriate in his case.

Then, during some of my courses on numerical analysis, I like to brandish the book in front of the students, and explain to them how awesome it is.