Objective To identify and characterize the loss of stereopsis observed in patients with lesions of the optic chiasm. Study design Cross-sectional study. Participants Forty-three patients who had good visual acuity with orthophoria and without strabismologic histories were divided into two groups. Group 1 consisted of 13 patients with lesions involving the optic chiasm (regardless of their visual field loss) diagnosed by magnetic resonance imaging findings. Group 2 (control group) consisted of 30 patients who had large absolute visual field defects as a result of other causes, including 11 intracranial disorders other than optic chiasmal lesions, 11 cases of open-angle glaucoma, and 8 patients with lesions of the retina. Methods The stereoacuity and visual field in each case in group 1 (before and after surgery) and group 2 were assessed, and the results were compared. Main outcome measures Stereoacuity was assessed by the Titmus stereo test (normal value for circle, 6/9; 80 seconds of arc) and by Lang-stereotest (normal value for circle, 3/3; 350 seconds of arc). Visual field was evaluated by Goldmann and Humphrey perimetry (conventional perimetry), the starlight test (binocular visual field test), and scanning laser ophthalmoscopic microperimetry (microperimetry). Results Before surgery, 11 of 13 cases (85%) in group 1 failed stereo tests, and after surgery, 5 of 13 cases (38%) in group 1 failed stereo tests. Before surgery, four patients who failed stereo tests showed no absolute scotoma by Humphrey or Goldmann analysis; after surgery, one patient who failed stereo tests showed no absolute scotoma by Humphrey or Goldmann analysis. However, starlight testing showed complete bitemporal hemianopsia only under binocular conditions, and microperimetry demonstrated a relative bitemporal hemianopsia at the fixating point. No patient failed in the Titmus circle test, but one patient in group 2 failed the Lang test (3%). The patients with chiasmal lesions significantly lost the ability of stereopsis compared with other diseases (group 1 [before or after surgery] vs. group 2, P < 0.001, Fisher’s exact test). Conventional perimetry was unable to measure scotomas within 3° of the fixation point, which is the key area for acute foveal stereopsis, because of an attached observational mirror. Conclusions The difficulty with stereopsis in patients with lesions of the optic chiasm is most likely caused by the compression of the decussating optic nerve fibers resulting in the loss of an overlapping visual field at the fixation point. Stereo tests were demonstrated to be simple and effective adjunctive tests for suspected chiasmal compression.