Objective Using a standard dichoptic noise-masking paradigm combined with qCSF method to assess the inter ocular suppression at different spatial frequencies in adults with amblyopia normal controls. Methods Ten adults with anisometropia amblyopia and 17 normal controls participated this research study. The degree of interocular suppression was quantified by the threshold elevation when the untested eye viewed a band-pass filtered noise compared to that when the untested eye viewed an unstructured field of the same mean luminance. Monocular signal attenuation by the amblyopic eye was accounted for by setting the contrast of the noise mask to a 5 times of its detection threshold. Data were analyzed using repeated measured ANOVA. Results This method was efficient in measuring the interocular suppression in both amblyopes and normal controls. We also found symmetric interocular suppression for normals (F=0.32, P＞0.05), but asymmetric interocular suppression for amblyopes (F=24.25, P＜0.01). In particular, the suppression from the amblyopic eye to the fellow eye was significantly less than that found in the normal controls (F=24.29, P<0.01), while the suppression from the fellow eye to the amblyopic eye was not significantly different to that found in the normal controls (F=3.81, P＞0.05). Conclusion Amblyopia involves an asymmetric interocular inhibition not accounted for by the signal attenuation of the amblyopic eye. This imbalance leads to dominance by the fellow eye over the amblyopic eye under binocular viewing (i.e. suppression).