Sinusoidal gratings of equal spatial frequency but different orientation require different levels of contrast to be detected by the human visual system. This phenomenon defined as oblique effect has a neuronal origin. The purpose of this work was to determine the neuronal magnitude of this effect, by isolating it from the optics of the eye. A visual interferometer was assembled to generate and project on the retina an interference pattern consisting of sinusoidal gratings with variable orientation (0º to 165º, 15º step). Adding background light to the interference pattern of 12 cycles/degree (cpd), different contrast levels were generated while the retinal illuminance was kept unaltered. A 2º circular stimulus was presented (during 500 ms) on the fovea producing a retinal illuminance of 134 Td (trolands). The contrast sensitivity threshold of four observers (ages 23, 33, 33, 52 years old) was determined using a Yes-No psychophysical method, and the 50% odds of correct response determined by a Weibull cumulative function. The four observers showed different contrast sensitivity thresholds dependent on the grating orientation. Oblique gratings (≈45º/≈135º) required more contrast to be detected than horizontal and vertical gratings. The maximum differences in contrast sensitivity between orientations ranged from 0.15 to 0.31 log units. The mean contrast threshold across all orientations was then calculated to investigate the effect of age on the contrast sensitivity. It was found a 0.046 log units decrease per decade (r=0.94). Oblique effect is an evident neuronal phenomenon with considerable inter-subject variability, making grating orientation important information in contrast sensitivity evaluation.