The article assesses the scale properties of the criterion set for severe sexual sadism in a sample of male forensic patients (N = 100). Half of the sample consists of sexual sadists; the remainder is sampled at random from the general group of nonsadistic sex offenders. Eleven of 17 criteria (plus the additional item of inserting objects into the victim's bodily orifices) of Marshall, Kennedy, Yates, and Serran's list form a cumulative scale. More specifically, this scale comprises all the 5 core criteria that Marshall and his colleagues considered particularly relevant. The resulting 11-item scale of severe sexual sadism is highly reliable (r(tt) = .93) and represents a strong scale (H = .83) of the Guttman type (coefficient of reproducibility = .97). The 11-item scale distinguishes perfectly between sexual sadists and nonsadistic sex offenders in the sample.