In the first part of the study, the effect of ipsilateral stimulus level on contralateral suppression of the 2f2-f2 distortion product by broad-band noise at 60 dB SPL RMS is examined. The levels of the primary stimuli were manipulated independently, giving f1 and f2 growth curves at four different f2 frequencies for four subjects. These typically bell-shaped, growth curves are shifted vertically to lower distortion levels and, in some cases, horizontally to higher primary stimulus levels. These results can be interpreted as an attenuation of both the primary stimuli and the distortion product and could be caused by simple acoustic attenuation produced by middle ear muscle activity, efferent activity or a combination of the two. In the second part of the study, the same contralateral stimulus was used while measuring both 3f1-2f2 and 2f1-f2 distortion products from the ipsilateral ear. The frequency separation of the primary tones was varied. This produced an approximately bandpass shape with the level peaking when the distortion frequency was approximately half an octave below f2, as previously described (Brown and Gaskill, 1990). This shape is thought to be linked with frequency selectivity in the cochlea. Contralateral broadband noise did not affect the tuning or the centre frequency of the bandpass shape or the mean group delay. It did reduce the size of the distortion peak and, in particular, it affected the peak-to-trough height of the 'fine structure' in the amplitude. Vector analysis revealed that the fine structure was due to a signal with substantial delay (probably from the distortion product 'place') which was summed with a larger, less delayed component (probably directly from the f2 'place'). The greater effect of contralateral stimulation on the more delayed component may reflect differences in efferent effect with complex (stimulus place), rather than simple (distortion product place) stimuli.