The aim of this study was to implement a familiarization protocol and to determine the constructive validity of four different non-reactive agility tests on a sample of professional female basketball players (n = 12; age: 18.17 ± 1.90 years). Following the protocol, the participants performed standardized tests three times (the Agility t-test and the Sprint 9-3-6-3-9 m with 180° Turns), while the lateral agility, universal agility, frontal agility, and semicircular agility tests were repeated five times each, with a three-minute rest period between each repetition. A single-factor analysis of variance for repeated measures (ANOVA) determined significant differences between a series of test repetitions for familiarization. Hierarchical cluster analysis grouped the samples based on the results of the standardized tests into two groups. A t-test for independent samples determined a significant difference between these groups (p < 0.05) on all four tests (lateral, universal, frontal, and semicircular agility). Pearson’s linear correlation determined a correlation between the results of the standardized agility tests and the lateral, universal, frontal, and semicircular agility tests, with correlation coefficients ranging from r = 0.936 to r = 0.987. A high value was observed for the discriminant analysis (0.821), while the maximum value was observed for the Kendall coefficient of concordance (W = 1). The study confirmed the validity of the applied tests and determined that the results are influenced by the number of series performed during the preparation for testing. In practice, it is necessary to perform two series of universal and frontal agility tests, i.e., four series of repetitions of lateral and semicircular agility tests in professional women’s basketball.