Recently, Combs and Holloman have reported anomalous energy minima in approximate SCF calculations on bending modes in several small molecules.1′2The methods employed (CNDO/2 and INDO)3are currently in wide use and it would be helpful to have a clear view of the origin(s) of these anomalies. [As has been demonstrated,2reparameterization to remove these features creates more serious problems than that for which the remedy was derived.] We were particularly concerned as recent studies have shown that INDO wavefunctions, when properly handled, provide a qualitatively correct description of the electronic charge distribution in selected diatomic molecules.4Further, these methods are extensively used for studies of molecular deformations.5