Purpose: We wanted to determine the mechanisms of motor vertical fusion in patients with superior oblique paresis and to correlate these mechanisms with surgical outcomes. Methods: Ten patients with superior oblique paresis underwent 3-axis, bilateral, scleral search coil eye movement recordings. Eye movements associated with fusion were analyzed. Results: Six patients had decompensated congenital superior oblique paresis and 4 had acquired superior oblique paresis. All patients with acquired superior oblique paresis relied predominantly on the vertical rectus muscles for motor fusion. Patients with congenital superior oblique paresis were less uniform in their mechanisms for motor fusion: 2 patients used predominantly the oblique muscles, 2 patients used predominantly the vertical recti, and 2 patients used predominantly the superior oblique in the hyperdeviated eye and the superior rectus in the hypodeviated eye. The last 2 patients developed the largest changes in torsional eye alignment relative to changes in vertical eye alignment and were the only patients to develop symptomatic surgical overcorrections. Conclusion: There are 3 different mechanisms for vertical fusion in individuals with superior oblique paresis, with the predominant mechanism being the vertical recti. A subset of patients with superior oblique paresis uses predominantly the superior oblique muscle in the hyperdeviated paretic eye and the superior rectus muscle in the fellow eye for fusion. This results in intorsion of both eyes, causing a large change in torsional alignment. The consequent cyclodisparity, in addition to the existing vertical deviation, may make fusion difficult. The differing patterns of vertical fusional vergence may have implications for surgical treatment. (J AAPOS 2002;6:145-53)