Fechner's paradox is a visual phenomenon which may be summarized as follows: a subject looks at an illuminated screen in binocular vision; one of the eyes is provided with a photometric neutral filter (density 1,3), the other eye remaining uncovered. When the uncovered eye is covered by means of a cover-test, the apparent luminosity of the viewing screen seems to decrease; but when the eye with the filter is covered by the cover-test, the apparent luminosity of the screen seems to increase: this fact is apparently paradoxical, since a part of the luminous stimulus of the eyes was suppressed: this is Fechner's paradox. The exact mechanism of this phenomenon is not yet clear; nevertheless, it is a primordial phenomenon of binocular vision, is easy to use in clinical strabological practice. Our present study was an application of Fechner's paradox in 119 strabismic subjects, 5 kinds of responses were obtained: 2 normal kinds of responses: Fechner's paradox symmetrical (41 subjects) or asymmetrical (29 subjects), i.e. 60% normal responses; and 3 kinds of pathological responses: mononuclear abolition (34 subjects), binocular abolition (10 subjects) and inversion of Fechner's paradox (5 subjects). The comparison between these responses and the clinical features was as follows: 1) Fechner's paradox was normal in 78% of subjects when the difference between the visual acuities of the two eyes was below 5/10, i.e. without true ambylopia; Fechner's paradox was abnormal in 74% of subjects when the difference was greater than 5/10; unilateral abolition of Fechner's paradox was not clearly related to the differences in visual acuity, but bilateral abolition and inversion of the phenomenon was related to ambylopia of one eye.(ABSTRACT TRUNCATED AT 250 WORDS)