In judo competition for visual impairment, athletes of different classes compete against each other in the same category; B1 athletes are totally blind, whereas B2 and B3 athletes are partially sighted. To test for potential competition disparities due a single category of athletes, this study aimed to compare competitive and technical–tactical performance in visually impaired judo athletes with different degrees of visual impairment. The authors analyzed 340 judo matches from the 2012 and 2016 Paralympic Games. The scores, penalties, efficiency index, and types of medals were examined, as well as the technical variation and temporal structure. The main finding was that blind judo athletes presented lower scores (p < .05; effect size [ES] = 0.43–0.73), medals (p < .05), and efficiency (p < .05; ES = 0.40–0.73); different patterns of play; and a shorter time to lose than partially sighted athletes (p = .027; ES = 0.10–0.14). However, the penalties were similar between classes (p > .05; ES = 0.07–0.14). The odds ratio of a winning medal was 3.5–8 times less in blind athletes than in partially sighted athletes (p < .01). In conclusion, blind judo athletes presented lower competitive and technical–tactical performance than athletes with some residual functional vision. These findings provide support for the development of new evidence-based criteria for judo classification based on vision impairment.