Computer based neurocognitive assessments are the most commonly utilized evaluative technique employed for the assessment of sport related concussion. Despite their wide acceptance among sports medicine professionals, the psychometric properties of these tests have yet to be established using clinically relevant testing intervals. PURPOSE: To establish the test-retest reliability of three commercially available computerized tests for concussion assessment. METHODS: One-hundred and eighteen healthy young adults (21.39 ± 2.78 years) completed the ImPACT, Concussion Sentinel, and the Headminder Concussion Resolution Index (CRI) tests on three occasions: Baseline, Day 45 and Day 50. Each participant also completed Green’s Memory and Concentration Test (MACT) to evaluate effort. RESULTS: Data from 73 participants were included in the analyses. Data were lost to poor Baseline test performance or failure to return. Intraclass correlation coefficient reliability estimates on each test from the initial day of testing to Day 45 were .15 to .39 on the ImPACT, .23 to.65 on the Concussion Sentinel, and .15 to .66 on the CRI. Slightly higher estimates were seen from Day 45 to Day 50 and were .39 to .61 on the ImPACT, .39 to .66 on the Concussion Sentinel, and .03 to .66 on the CRI. High effort was exerted by all participants as indicated by results on the MACT. CONCLUSION: These findings suggest that the test-retest reliability of commonly utilized computer-based tests for concussion are less than optimal when administered using a test-retest interval relevant to the sports medicine practitioner. Inconsistent performance on these tests may lead the clinician to inaccurately evaluate neurocognitive performance following a suspected concussion. Those implementing a concussion assessment protocol for athletes at risk for concussion should consider combining the neurocognitive assessment with other evaluative tools known to be sensitive to the effects of concussion.