Tested whether there is a specialized speed processing mechanism for stereo-defined motion. Speed discrimination thresholds were compared for stimuli containing only stereo-defined motion, and stimuli that contained both stereo-defined and luminance-defined motion with 2 Ss. When the stimulus contained luminance-defined motion, increment thresholds were around 0.05–0.1. For stereo- defined motion, increment thresholds were never better than 0.3. By careful analysis, it was possible to test what cues were being used to solve the speed discrimination task. Results are consistent with observers responding to distance cues rather than to speed for stereo-defined motion, suggesting that there is no specialized mechanism for processing the speed of stereo-defined motion.