A global three-dimensional chemical transport model has been used to simulate atmospheric bromoform using a variety of prescribed surface emission scenarios and a simple atmospheric chemistry scheme. Model simulations indicate that global emissions of bromoform calculated previously using top-down methods are too low, and emissions are likely to be significantly larger than suggested in the World Meteorological Organization's reports on the Scientific Assessment of Ozone Depletion of 1998 and 2002. Our simulations suggest that global emissions of bromoform are in the range of 400--600 GgCHBr3/yr and that a large proportion of emissions are situated in tropical regions. Both these factors are likely to have an important influence on estimates of the quantity of bromine transported from the surface to the lower stratosphere by short-lived bromocarbon species and its subsequent impact on ozone in this region. Further simulations including methyl bromide, bromoform and the other major short-lived bromocarbons emitted from the ocean (CH2Br2, CH2BrCl, CHBr2Cl, CHBrCl2) provide estimates on the amount of reactive bromine in the troposphere and lower stratosphere derived from these compounds.