We examined anisotropy of filling-in in the blind spot in horizontal and vertical orientation. A pair of lines was presented across the blind spot with varying lengths of both line in experiment 1 and varying length of a single line in experiment 2, the observers reported whether the line appeared‘complete’or‘gapped’. In experiment 3, a pair of patterns, which were consisted of 3, 7, or 11 lines, was presented across the blind spot, and observers were required to report whether or not the patterns were perceived continuous. The results showed that the minimum length of line for filling-in was shorter in horizontal orientation than that in vertical, and that filling-in occurred more in horizontal lines than in vertical. We discussed the obtained anisotropy with respect to the spatial resolution, the cortical magnification factor, and the shapes of receptive fields of neurons that caused filling-in in the blind spot.