A critique of the recently proposed activation-synthesis hypothesis on the origin and formation of dreams is presented. Many findings do not support the new hypothesis that specific pontine physiological processes, rather than mental processes, instigate dreams and produce their distortion. First, dreaming often occurs in the absence of the pontine processes. Second, forebrain activity (which can have mental correlates) is crucial to the instigation and maintenance of dreaming sleep. Finally, activity of the proposed pontine dream generator, which is claimed to cause dream distortion, is not reliably accompanied by dream distortion.