To the Editor.—The article "Radiation-Induced Parotid Cancer" by Walker et al (Archives 1981;116:329-331) states erroneously that the incidence of radiation-induced parotid cancer is 8%. The information presented is that six of 72 patients who received treatment at the authors' hospital for various parotid cancers, and included in their analysis, had a history of prior treatment with radiation, much as 100% had a history of eating bread. Any study of the incidence of "radiation-induced cancer" is meaningless without a denominator (number of patients receiving parotid irradiation). This incidence could then be compared with the incidence in a control (unirradiated) population. Also, the probability of radiation induction of cancer is a complex function of dose1; it is unwarranted, on the basis of the authors' data, to claim that a dose of 300 rad has any significance as a threshold.