Almost all differential equation in physics can be derived from a suitable Lagrangian using variational methods. I was wondering, is there a general theorem that says that every differential equation has its "Lagrangian"? If yes, then Lagrangian approach to physics is tautological.