The purpose of this study was to determine the efficacy of computerized fusional vergence therapy and the effect of two different vergence training velocities. Six subjects received positive vergence training using a slow vergence training rate (0.75 < sup > Δ < /sup > /s) and six subjects received positive vergence training using a fast vergence training rate (5.00 < sup > Δ < /sup > /s). Six subjects served as controls and did not receive therapy. The duration of therapy was 80 min over a period of 4 weeks. All training activities were monitored. All vergence evaluations were double masked. Subjects using a slow training rate showed significant in-creases in positive vergence ranges as measured with the major amblyoscope, whereas subjects training with fast rates did not. We conclude that vergence therapy using a computerized video display is an effective technique for increasing the amplitudes of positive fusional vergence and that slower rates are more productive than faster rates. © 1987 American Academy of Optometry.