Our sample of antimony oxide had a low acute and chronic toxicity for rats, probably because of slight solubility. It appears to have been considerably less toxic and therefore less soluble than the sample studied by Bradley and Frederick. It is not unusual for different samples of an oxide to differ appreciably in solubility due to differences in crystal structure but no x-ray diffraction measurements were made to define the structure of our sample. If antimony oxide acts as a stabilizer for vinyl resins by accepting hydrochloric acid to form antimony trichloride, it seems likely that feeding tests on the oxide may be inadequate to judge the safety of its use in contact with foods. A hydrated oxide or oxychloride formed by contact of the trichloride with water might be considerably more soluble and therefore more toxic than the oxide as it is added to the film in manufacture. Nevertheless our screening tests with antimony oxide are sufficiently promising to justify further study of the stabilizer.