To investigate whether constant exotropia patients with a previous history of intermittent exotropia X(T), like X(T) patients, can achieve postoperative bifixation, and whether they have a better postoperative sensory outcome than those without previous history of X(T). Prospective comparative clinical study. Sixty-three consecutive patients with intermittent or constant exotropia were divided into three groups: X(T) (group 1), constant exotropia with a previous history of X(T) (group 2), and constant exotropia without previous history of X(T) (group 3). The surgical outcomes were assessed and compared in motor and sensory terms separately between the three groups. Successful motor alignment was defined as within 8 prism diopters (PD) (exo or eso). A stereoacuity < or =60 seconds of arc was considered as bifixation, and a stereoacuity < or =800 seconds of arc was considered as gross stereopsis. The successful motor alignment rates of group 1, group 2, and group 3 were 79%, 71%, and 67%, respectively (group 1 vs group 2, P = .826; group 1 vs group 3, P = .551; group 2 vs group 3, P = 1.000). Twenty-five (74%) patients in group 1 achieved bifixation and none achieved in group 2 or group 3 (group 1 vs group 2, P = .001; group 1 vs group 3, P = .001). Meanwhile, 34 patients (100%) in group 1, 11 (79%) in group 2, and 5 (33%) in group 3 achieved gross stereopsis (group 1 vs group 2, P = .021; group 2 vs group 3, P = .025; group 1 vs group 3, P = .001). Compared with patients in the two constant exotropia groups, patients in X(T) group had a significantly better sensory outcome in both bifixation and gross stereopsis. Patients in group 2 had a better sensory outcome than those in group 3 in gross stereopsis. Constant exotropia patients with a previous history of X(T) have a better postoperative sensory outcome in gross stereopsis than those without previous history of X(T), but a worse surgical sensory outcome when compared with X(T) patients in both bifixation and gross stereopsis. Constant exotropia patients decompensated from X(T) may have missed the best time for treatment.