Field experiments were conducted at Warooka and Rudall on the Yorke and Eyre Peninsula of South Australia during 2004 and 2005 to investigate the effectiveness of metribuzin for the selective control of rigid brome in no-till sown barley. Metribuzin (135 to 203 g ai/ha) incorporated by sowing (IBS) was more effective in controlling rigid brome (> 67%) than the same herbicide dose applied POST. Although IBS metribuzin at the highest rate (270 g/ha) provided effective weed control (82 to 90%), it was more phytotoxic to barley, reducing crop density by 23% relative to the nontreated control at Rudall. Soil at Rudall had low clay and organic matter content. In contrast, tank mixtures of metribuzin (203 g/ha) with pendimethalin, applied IBS, provided reliable rigid brome control (89 to 93%) and resulted in little crop damage (< 5%). Over the site-by-year combinations studied, all herbicide-treated barley at Warooka and Rudall yielded 6 to 50% more grain than the nontreated crop. Although metribuzin provides growers with an opportunity to selectively control rigid brome in no-till barley, high rates (≥ 203 g/ha) of this herbicide on sandy textured soils can result in significant crop damage.