ABSTRACT Purpose: This study investigated the test-retest reliability and discriminant validity of the Kickboxing Anaerobic Speed Test (KAST) and established a comparison with the maximal cycling sprint test (MCST). Methods: Forty-two male kickboxing athletes (18 elite and 24 sub-elite) participated in this study. Kickboxers performed the tests in both single and multiple forms. The KASTsingle was composed of kicks and punches, performed as quickly as possible until 5 reps of a 4 techniques’ combination (i.e., straight left punch, right-roundhouse kick, straight right punch, and left-roundhouse kick) were completed. The MCSTsingle was composed by a single 6-s maximal effort. The KASTmultiple was performed using 5 sequential sets of the single form following a 10s rest interval between sets. The MCSTmultiple was performed using 5 × 6 s repeated cycling efforts with 10s rest intervals. The tests and retests were carried out on separate occasions. To establish the test’s discriminatory capability, elite and sub-elite athletes were compared. Results: All tests’ performance scores showed excellent relative and absolute reliability (ICC > 0.900, SEM ≤ 0.98 s for KASTs and ≤ 0.74 W•kg−1 for MCST tests). Significant correlations between the identical versions were “large” (r > 0.70). Receiving operating characteristic analyses indicated that the KASTs and cycling tests were able to effectively discriminate between elite and sub-elite kickboxers. Additionally, the findings showed that KASTsingle, MCSTsingle, KASTmultiple, and MCSTmultiple correctly classified the groups by 78.6%, 73.6%, 88.1%, and 78.6%, respectively. Conclusions: This study supported the test-retest reliability and the discriminant validity of the KASTsingle and KASTmultiple to evaluate kickboxing athletes.