The main purpose of this study was to determine the relationships among sprinting performance, change of direction ability (COD), change of direction deficit (CODD), and aerobic power expressed by maximal oxygen velocity (VmaxIFT), anaerobic speed reserve (ASR) as well as jumping performance (countermovement jump with (CMJa) and without an arm swing (CMJ)) in elite 3x3 basketball players. A total of 15 Polish Olympic 3x3 team players (age: 26.86 ± 8.28 years; body height: 191 ± 5.33 cm; body mass: 90.68 ± 10.03 kg, basketball experience: 15.53 ± 5.8 years) participated in the study. Athletes were tested for the following measures: the linear speed at the first section (5 m), the second section (9 m), the third section (10 m) and total distance (24 m), two sets; the 30-15 Intermittent Fitness Test (first session); COD speed by the Change of Direction and Acceleration Test (CODAT) (the same sections and total distance as in the linear speed test), five repetitions, two sets; and jumping performance by the CMJ with and without an arm swing (second session). CODD was calculated by subtracting the COD speed time from linear speed time at adequate sections and total distance. Maximal sprinting speed (MSS), maximal aerobic velocity (VmaxIFT), and anaerobic speed reserve (ASR) were also considered. Anaerobic Speed Reserve (ASR) was calculated as the difference between MSS and VmaxIFT. A Pearson’s correlation test was used to determine the relationship between power-speed-related variables and CODD, final velocity attained at the 30-15 Intermittent Fitness Test (VmaxIFT), ASR, and COD performance. Moderate to strong correlations were registered between COD and linear speed at 5, 10, and 24 m, while moderate to strong negative correlations were detected between COD, CODD, and CMJ, CMJa. Moreover, moderate to strong correlations were observed between COD, CODD, and VmaxIFT, MSS at 9, 10, and 24 m sprints. No relationship was detected between COD, CODD, LS, and ASR in any measured sector. Finally, statistically significant differences were registered in COD and CODD between trial 1 and trial 2.