The widely held belief that Hartree-Fock orbitals 1 behave asymptotically like exp[-(-2εi)12r], where εi is the orbital energy of i, is shown to be incorrect, with one exception: atomic configurations consisting entirely of s orbitals. The correct asymptotic form of i is a sum of terms like exp[-(-2×εj)12r], in which all εj appear. The former misconception apparently resulted from too superficial a treatment of the exchange potential at large r. © 1969 The American Physical Society.