Factors influencing dicamba drift, especially vapor drift, were examined in field and growth chamber studies. In field experiments, potted soybeans [Glycine max (L.) Merr.]. exposed to vapors arising from corn (Zea mays L.) foliarly treated with the sodium (Na), dimethylamine (DMA), diethanolamine (DEOA), or N-tallow-N,N1,N1-trimethyl-1,3-diaminopropane (TA) salts of dicamba (3,6-dichloro-o-anisic acid), developed dicamba injury symptoms. Dicamba volatilization from treated corn was detected with soybeans for 3 days after the application. Dicamba vapors caused symptoms on soybeans placed up to 60m downwind of the treated corn. When vapor and/or spray drift caused soybean terminal bud kill, yields were reduced. In growth chamber studies, dicamba volatility effects on soybeans could be reduced by lowering the temperature or increasing the relative humidity. Rainfall of 1mm or more on treated corn ended dicamba volatilization. The dicamba volatilization was greater from corn and soybean leaves than from velvetleaf (Abutilon theophrasti Medic.) leaves and blotter paper. The volatilization of dicamba formulations varied in growth chamber comparisons with the acid being most volatile and the inorganic salts being the least volatile. However, under field conditions, use of less volatile formulations did not eliminate dicamba symptoms on soybeans. The volatile component of the commercial DMA salt of dicamba was identified by gas chromatography-mass spectrometry as free dicamba acid.