Full canopy closure and light interception are critical to obtaining full yield potential of ultra-short–season soybean in the midsouthern United States. We hypothesized that herbicide applications that resulted in soybean leaf injury would reduce season-long light interception and yield of ultra-short–season soybean grown in this environment. Experiments were conducted in 2001, 2002, and 2003 at Fayetteville, AR, to determine the effect of the diphenylether herbicides acifluorfen and lactofen on light interception and yield of maturity group (MG) 0 and II soybean. Factors evaluated included soybean MG, herbicide rate, treatment timing, and soybean seeding density. When applied at soybean growth stage (GS) V3, 0.2 kg ai/ha lactofen reduced green leaf area immediately after application and final canopy closure relative to soybean treated with 0.6 or 0.2 kg/ha acifluorfen and untreated soybean. Herbicide application did not affect yield of well-watered soybean when applied at GS V3 in 2001 or at early reproductive development in 2003. In 2002, an irrigation problem resulted in a period of water-deficit stress during seed fill of MG II soybean. Under these conditions, treatment with acifluorfen at GS V3 reduced soybean yield, and treatment with lactofen during early reproductive development reduced soybean yield, relative to untreated soybean. This research indicates that diphenylether herbicides can be safely applied to well-watered ultra-short–season soybean, but yield reduction can occur when applied to soybean that is not well watered.