The purpose of this study was to evaluate the reliability and validity of a new test of agility, the reactive agility test (RAT), which included anticipation and decision-making components in response to the movements of a tester. Thirty-eight Australian football players took part in the study, categorized into either a higher performance group (HPG) (n=24) or lower performance group (LPG) (n=14) based on playing level from the previous season. All participants undertook testing of a 10m straight sprint (10mSS), a 8-9m change of direction speed test (CODST), and the RAT. Test-retest and inter-tester reliability testing measures were conducted with the LPG. The intra-class correlation (ICC) of the RAT was 0.870, with no significant (p<0.05) difference between the test results obtained on the first and second test sessions using a t-test. A dependent samples t-test revealed no significant (p<0.05) difference between the test results of two different testers with the same population. The HPG were significantly (p=0.001) superior to those of the LPG on the RAT, with no differences observed on any other variable. The RAT is an acceptably reliable test when considering both test-retest reliability, as well as inter-rater reliability. In addition, the test was valid in distinguishing between players of differing performance level in Australian football, while the 10mSS and CODST were not. This result suggests that traditional closed skill sprint and sprint with direction change tests may not adequately distinguish between players of different levels of competition in Australian football.