One of the most widely used techniques for suppression of solvent resonances in 1H FT NMR spectra is the WATERGATE method (water suppression by gradient tailored excitation). Excellent suppression of a water resonance can be achieved without the concomitant reduction of resonances that occurs when, for instance, the presaturation method is used. Very poor suppression is experienced when WATERGATE is used with the DQF-COSY method (and also short mixing time NOESY and TOCSY experiments). This is thought to be caused by the absence of a delay in which radiation damping can assist in restoring water magnetization back to the +z axis prior to the read pulse. If, however, the WATERGATE segment is applied after the initial excitation pulse (rather than after the read pulse as is usually done for WATERGATE experiments), then good water suppression is obtained. In this poster the effects of moving the WATERGATE segment from the end to the beginning of the pulse sequence is examined, and results compared with other methods for obtaining DQF-COSY spectra in aqueous solutions.