Change-of-direction ability is an important performance skill in football. Consequently, several testing protocols are used to assess this component. This study assessed the test-retest reliability of a modified 505 test (M505) and the change-of-direction deficit (CODD) in elite youth football players. Data were collected from 110 players from the Under [u] 12-18 years age groups (u18 n = 26, u16 n=26, u14 n=39, u12 n=19) within two English Premier League Category 1 Football Academies. Players completed the M505 twice in 7-days, in addition to a 10-m sprint test to allow CODD to be calculated. Reliability was assessed with respect to chronological and biological age (according to PHV status). Typical error (%), minimal detectable change (MDC%) and intraclass correlation coefficients (ICC) were calculated. Typical errors in M505 were moderate to large (2.0 to 3.2%), while intraclass correlation coefficients (ICC) ranged from low to high (r = 0.26 to r = 0.82). Typical errors in CODD were moderate to large (7.1 to 12.0%), with ICC’s ranging from low to high (r = 0.19 to 0.79). Minimal detectable changes were 5.5 to 8.9% in M505 and 17.7 to 33.3% in CODD. The typical errors and minimal detectable changes observed here indicate that the M505 and CODD tests have limited practical utility in the evaluation of change of direction ability in elite youth football players.