The importance of having attenuation curves available of radiation scattered from objects in the useful beam, such as the walls of a radiation room or a patient, has already been pointed out (1). Each irradiated object becomes a source of secondary radiations which may be hazardous to personnel located in areas away from the direction of the useful beam. Only by knowing the attenuation curves of the secondary radiation in common protective materials, such as lead or concrete, can the most economical secondary protective barriers be designed to eliminate this hazard. The data available on the attenuation of scattered gamma rays are of two types. For one a suitable phantom is used to duplicate the conditions under which a patient scatters radiation out of the useful beam (2, 3). The second type consists of measurements in which the radiation is scattered from a concrete wall or barrier (2). The scattered radiation may traverse a longer path within the phantom than in the wall and thus be more greatly attenuated, so that the dose rate at a given distance from the scatterer may be higher for the second type of measurement. For a given moderate-to-hard quality of useful beam radiation, however, the quality of scatter at a given angle should not be very different for the two cases. Although interest in teletherapy sources has been directed chiefly to Co60 (2, 3) up to the present time, several other isotopes have characteristics which make them desirable for this purpose. Brucer (4) has pointed out the requirements of teletherapy source materials: source energy between 0.2 and 1.0 Mev to reduce the shielding requirements, high specific activity for small source size, long life, and availability. Cs137 with its thirty-year half-life and an energy of 0.66 Mev fulfills these requirements better than C0 60 with a 5.2-year half-life and an energy of 1.2 Mev. Experimental Procedure The geometry used in the experiment is shown in Figure 1. The 120-curie Cs137 source consisted of CsC1 compressed into a stainless steel can with 1-mm. walls within a lead container approximately 11 in. in diameter, having a horizontal-axis conical opening of half angle 8.5°. The vertical scattering barrier, made up of concrete blocks 6 · 8 · 12 in. laid up into a wa1140 in. wide, 42 in. high, and 8 in. thick, standing on a table 24 in. high, was located so that the source-to-barrier distance measured by the beam axis was 100 cm. and the barrier was at an angle, β, with respect to the axis of the incident gamma rays. The barrier was thick enough so that any increase in thickness would yield negligible increase in the scattered dose rate, as determined by an auxiliary experiment.