Abstract
The modification of Levinson's theorem for the case of non-local potentials suggested by many authors is shown to be incorrect. The error is a result of overlooking a simple property of the inverse tangent function.