The rapid adoption of dicamba (3,6‐dichloro‐2‐methoxybenzoic acid)‐resistant (DR) soybean [Glycine max (L.) Merr.] resulted in an increase of post‐emergent dicamba applications during the soybean‐growing season, resulting in off‐target movement and injury to non‐DR soybean. Field trials were established in Manhattan, KS, in 2018 and 2019 and in Ottawa, KS, in 2019 to characterize the response of non‐DR soybean to one, two, or three applications of reduced rates of dicamba at three application timings. Soybean were treated with 0.56, 1.12, and 5.6 g a.e. ha−1 of dicamba, which is equivalent to 1/1,000X, 1/500X, and 1/100X of a 1X field‐use rate (560 g a.e. ha−1), respectively. Soybean plants were treated at V3, R1, R3, V3 followed by (fb) R1, V3 fb R3, R1 fb R3, and V3 fb R1 fb R3 growth stages. Soybean injury from dicamba was less severe following application during the V3 than the R1 or R3 growth stages. In general soybean injury was the greatest 4 wk after application. The greatest soybean yield reduction (68%) followed dicamba applications of 5.6 g a.e. ha−1 at V3 fb R1 fb R3 in Manhattan, KS, 2018, where yield loss was generally greater and may be attributed to droughty conditions. Yield loss was minimal in Manhattan, KS, and Ottawa, KS, in 2019 following a single dicamba application at the V3 stage, regardless of application rate and following dicamba application at 0.56 g a.e. ha−1, regardless of number of applications. The greatest soybean yield losses from dicamba occurred with two or three applications at 1.12 or 5.6 g a.e. ha−1.