Observers with normal stereopsis suppress some of the monocular information contained in each stereo half-image, a phenomenon we call fusional suppression. We measured vernier acuity for an ordinary vertical vernier test target, presented to one eye, that was paired stereoscopically with a vernier target with a large fixed offset, presented to the other eye. The size of the fixed offset, and hence, the disparity of the upper target line, was varied parametrically. Vernier thresholds for the test target rose as a function of disparity, reaching a maximum at 20 min of disparity and then decreasing gradually as the disparity exceeded the limits of foveal fusion (40-60 min arc). In the two normal observers, fusional suppression was symmetrical; neither eye had good access to monocular information. In the stereoanomalous observers, fusional suppression was not symmetrical. When they viewed the vernier test target with the stronger of their two eyes, their vernier thresholds were barely affected by the stereo half-image in the other eye, and so were better than those of the normal observers measured in the same condition. When the stereoanomalous observers viewed the test target with their weaker eye, their fusional suppression was similar in range and magnitude to the suppression found in normal observers. Amblyopic suppression in mild amblyopes may be a residual effect of normal fusional suppression, operating to suppress monocular signals in the weaker eye, without conferring the benefits of normal stereopsis and fusion.