e have read Foote et al.' with interest. Certainly a W modern series showing the efficacy of 6 megavolt (MY) photon therapy in the treatment of early glottic cancer is reassuring but it is important to note the effect of fraction size. Updated data regarding treatment with 6 MV linear accelerators2 reveals that 22 T1 glottic cancer patients have now been treated with 2.25 gray (Gy) fractions to 63 Gy with 95% 2-year local control (Foote RL, personal communication, 1996). Isn't it possible that the prescription dose per fraction of 2.25 Gy compensated for the decreased dose at the tumor surface as a consequence of the higher energy linear accelerator? If so, Foote et al. have confirmed Schwaibold et al.'s analysis of T1 glottic cancer,' which found that age, sex, hemoglobin, tumor bulk, differentiation, field size, total dose, and total treatment time were not significant predictors of local control, whereas fraction size of 2 Gy or greater was critical to local control. Of course the 4 MV linear accelerator used by Schwaibold et al. would have a dose distribution similar to the bi-energetic Co-60.The "magic" dose per fraction figure may be the dose at the free edge or anterior commissure. Perhaps we can use any of a number of methods of delivering the dose as long as we "overprescribe" to ensure the free edge and anterior commissure are treated to the critical daily tumoricidal dose.Why should daily fraction size matter if the overall treatment time does not? It has been pointed out3 that the lack of an effect of overall treatment time on outcome may be due to the shape of the sigmoid dose response curve, and prolongation of the treatment time will be most important when the tumor control probability is steeply rising.The dependence of local control on fraction size is the best evidence proliferation is alive and important for early glottic cancer. Since demonstrable improvement over current results for T1 glottic cancer is difficult, there is little incentive to further increase the fraction size. Conversely, if the critical dosimetry is within mm of the tumor surface underdosage at the mucosal surface of exophytic tonsillar or nasopharyngeal tumors might also be detrimental if treated with 2 6 MV linear accelerators. Rouby E, et al. The effect of fraction size on control of early glottic cancer.