A 'random walk' visual training system was developed to be presented on a haploscope. This test entailed a series of 2 and 4 deg step changes located at different vergence angles presented for an unpredictable amount of time. This training sequence was designed to serve as a control measurement for future comparison to evaluate whether three dimensional gaming stimulates adaptation to the vergence system. The random walk haploscope training is core of this paper. Five binocular normal controls participated in 12 hours of training consisting of visual tracking stimuli at near vergence angles (16, 18, 20 degrees), at far vergence angles (1, 3, 5 degrees), and smooth visual tracking (ramps) from near to far stimulating different tracking speeds. Significant improvements were not observed in peak velocity and the time to peak velocity for most subjects. The subject within the group studied who exhibited the longest reaction time to attain peak vergence velocity did exhibit a significant decrease in the amount of time to attain 90% of the steady state after 12 sessions of vergence training. Results suggest that subjects who have longer reaction times may improve their ability to fuse new targets or improve precision after repetitive vergence training.