The aim of this study was to investigate the relationship between linear sprint, power output obtained during a squat and change of direction (COD) performance. Fifteen elite soccer players participated in this study (age = 21.7 ± 0.72 years, body mass = 74.9 ± 9.11 kg, body height = 180.4 ± 7 cm, training experience = 9 ± 1.5 years). To examine these correlations a following battery of tests were carried out: 20-m linear sprint, one-repetition maximum (1RM) squat strength, peak power output obtained during a squat at 50% 1RM and time obtained in two 20-m COD tests with different angles of direction change (90° and 135°). In addition, COD deficits (90°-CODDEF and 135°-CODDEF) for both COD tests were calculated. The Spearman’s rank order correlation showed a nearly perfect statistical relationship between the 90°-COD and the 90°-CODDEF (r = 0.9; p < 0.001). In the case of 90°-CODDEF, there was a large statistical relationship with 135°-CODDEF (r = 0.59; p = 0.021). Moreover, there was a nearly perfect statistical relationship between 135°-COD and 135°-CODDEF (r = 0.91; p < 0.001). The statistically insignificant (p > 0.05) relationship between 20-m linear sprint time, power output obtained during a squat at 50% 1RM, 1RM squat strength level and both COD test, as well as both COD deficits were found. Results of the present study showed that 20-m linear sprinting speed, 1RM squat strength, power output obtained during squat at 50% 1RM and COD ability at 90° and 135° angles, are separate physical qualities. Moreover, it seems that COD deficit provides a more isolated measure of COD ability than the COD tests alone and does not must be limited to a specific angle, but provides knowledge about the COD ability in a range of other angles, at least concerning 90° and 135° COD angles.