Field studies were conducted in Georgia, North Carolina, South Carolina, Tennessee, and Virginia during 2006 to investigate tolerance of cotton to fomesafen applied preemergence (PRE). Fomesafen at seven rates, and two standard herbicides, pyrithiobac and fluometuron, were applied PRE to cotton in a weed-free environment. Cotton tolerance to fomesafen was directly related to rainfall that occurred from planting through cotton emergence. No injury was detected in South Carolina or Tennessee, but heavy rainfalls prior to cotton emergence in Georgia led to 3 to 9% early season visible stunting by fomesafen at 140 to 420 g a.i. ha-1 and 11 to 15% stunting by fomesafen at 560 and 840 g ha-1. In North Carolina and Virginia, rainfall during cotton emergence led to early season cotton necrosis ranging from 4 to 16% with fomesafen at 140 to 350 g ha-1 and 12 to 45% by fomesafen at 420 to 840 g ha-1. Early season injury by fomesafen at 280 g ha-1 (recommended use rate) was equal to or less than pyrithiobac or fluometuron at 4 of 5 locations. Mid-season injury was 10% or less at all locations with fomesafen at 490 g ha-1 or less. Plant heights were reduced 11 to 29% in Georgia and North Carolina when fomesafen was applied at rates greater than 420 g ha-1. In Tennessee, heights were reduced 8% with fomesafen at 560 g ha-1. Compared to the non-treated control, plant stands were reduced 23 to 28% only in North Carolina when fomesafen rates exceeded 350 g ha-1. Lint yields followed trends in plant stand, with yield being reduced 23 to 25% by fomesafen at 560 to 840 g ha-1 only in North Carolina. Cotton fiber quality and cotton fruit distribution or number of fruit set, were not adversely affected by herbicides.