Bovine zygotes produced by in vitro oocyte maturation and fertilization were cultured for 7.5 d in a chemically defined medium without serum or proteins, except .12 IU/mL of insulin. In Exp. 1, embryos were cultured in approximately 20% oxygen (i.e., 5% CO₂ in air) or 5% CO₂; 5% O; 90% N₂, with the metal chelators EDTA or diethylenetetraaminopentaacetic acid (DTPA) at 0, 5, 25, or 125 µM. More (P < .01) embryos developed to blastocysts at 5% O₂ (17%) than at ~20% O₂ (7%). Also, embryos grown at 5% O₂ averaged more cells than embryos cultured at ~20% O₂ (38 vs 29 cells for morulae and blastocysts and 15 vs 12 cells including all embryos; P < .05). There were interactions (P < .01) among chelator, concentration of chelator, and oxygen tension. The most efficacious treatments were 5 µM EDTA at 5 or ~20% O₂ (24 and 20% blastocysts), 5 µM DTPA at 5% O_` (28% blastocysts), and 25 µM EDTA at 5% O₂ (25% blastocysts). High concentrations of either chelator were detrimental, especially at ~20% O₂. In Exp. 2, a smaller range of chelator concentrations was compared (EDTA: 3, 9, 27, or 81 µM, DTPA: 3 or 15 µM) in 5% O₂. More embryos developed to blastocysts and expanded blastocysts with 3 µM EDTA than the control without a chelator (20 and 16% vs 7 and 3%, respectively; P < .05). However, in Exp. 3, which concerned embryo development in .33, 1, 3, or 27 µM EDTA and .33, 1, or 3 µM DTPA, no concentration of either chelator was better (P > .1) than the control.