Grain yield and its endpoint components were studied for three barley varieties, Dayton, Hudson, and Tenn 60-34, as pure stands and two-way and three-way mixtures at three widely separated sites for two years. The populations, including the pure stands, produced similar grain yields. Complementary competition was observed in all mixed populations in which Tenn 60-34 occurred. Its yield was suppressed in mixtures about 20% while Dayton and Hudson increased a like amount so that total mixed population yield was equal to that expected from pure stand performance. Tenn 60-34 occurred at 33, 50, and 80% frequencies in mixed populations with no difference in competition, indicating a lack of frequency dependency for the effect.