This study aimed to test the reliability, sensitivity, construct and logical validity of an adapted Special Judo Fitness Test (SJFT) for judo athletes with visual impairment . Twenty judo athletes with visual impairments performed both the adapted SJFT with tactile and sonorous stimuli (experimental conditions) and the typically administered SJFT (standard condition). We used analyses of variance (ANOVAs) with repeated-measures to compare the groups’ SJFT performances, and one-way ANOVAs to compare different visual ability classes of athletes (B1, B2 and B3). We used t-tests to compare SJFT variables between elite and sub-elite groups. We set statistical significance for all tests at p < 0.05. The standard SJFT showed excellent test-retest reliability for number of throws and overall index (ICC = 0.91–0.95), and both sonorous and tactile sensitivity adaptations of the SJFT showed medium sensitivity for detecting performance changes. The number of throws and SJFT index were higher with the sonorous adaptation of the test, compared to the tactile and standard versions ( p < 0.001). Athletes who were blind (B1) presented similar performances to athletes who were partially sighted (B2 and B3) only on the SJFT with the sonorous stimulus. Moreover, only the sonorous SJFT adaptation discriminated between the performances of elite and sub-elite athletes ( p < 0.001). In conclusion, both SJFT adaptations showed excellent reliability and medium sensitivity on test-retest, but, only the SJFT with the sonorous stimulus seemed valid for assessing judo athletes with varying degrees of visual impairment, and only the sonorous stimulus SJFT discriminated elite from sub-elite athletes.