Abstract
The author gives a critical analysis of a derivation due to Kubo (1966) of the fluctuation-dissipation theorem for classical systems. It is shown that a basic assumption having the appearance of a causality condition actually should be understood as an incidental corollary of the equations.