BackgroundWe prospectively evaluated the performance of three stereoacuity tests in patients with a wide range of strabismus angles to determine the maximum angle of horizontal strabismus consistent with true stereopsis as well as the extent of false-positive results. MethodsSubjects age ≥ 5 years old (n = 186) with constant horizontal strabismus were prospectively studied. Stereoacuity was measured using the Titmus Fly, Animals, and Circles tests; the original Preschool Randot test; and the Frisby test. We defined a false-positive result as any positive response to a stereoacuity test when a patient had > 20 prism diopters (PD) of deviation. Based on the results from our previous study of monocularly patched adults, we then defined a true-positive result as Titmus circles five to nine, at least 400“ on the original Preschool Randot test, or a pass on the Frisby test. Data were then analyzed to determine a maximum deviation consistent with true stereopsis. ResultsFalse-positive results occurred with Titmus Fly (6%), Titmus Animals (10%), Titmus Circles (35%), and the 800 seconds of arc (”) level of the original Preschool Randot test (10%). No false-positive results were seen with the Frisby test. Applying the criteria for a true-positive result, we found that no patient with > 10 PD of deviation had true stereopsis. The two patients with 10 PD and 6 PD deviations could not subsequently pass the Frisby test and were therefore deemed not to have true stereopsis. We concluded that the threshold for true stereopsis may be as low as 4 PD of horizontal deviation. ConclusionIn summary, the Titmus Fly, Titmus Animals, and Titmus Circles (the first four circles) tests possess monocular clues that limit their usefulness for clinical testing. The Frisby test is particularly useful for rapid assessment of whether stereopsis is present or absent. The new Preschool Randot test is valuable for quantifying stereopsis in both children and adults. True stereopsis may be rare when a patient has a horizontal deviation > 4 PD.