The purpose of the present study was to investigate the intersession reliability of vertical jump height in women and men recorded from a contact mat. Thirty-five women and 35 men performed four testing sessions across a 4-week period, with each session separated by 1 week. Within each testing session, subjects completed three countermovement vertical jumps (CMJs) for maximum height. Reliability statistics were calculated using the highest jump (HIGH) and also from the mean of all three jumps (3 MEAN) during each session. Reliability was calculated as a change in the mean, coefficients of variation (CVs), and intraclass correlations coefficients (ICCs) between testing sessions. For women, jump heights were not substantially different between sessions for either the HIGH or 3 MEAN data. The CVs for women ranged from 4.4 to 6.6% for HIGH and 4.1 to 6.0% for 3 MEAN, with the corresponding ICCs ranging from 0.87 to 0.94 for HIGH and 0.90 to 0.95 for 3 MEAN. For men, jump heights were not substantially different between sessions for HIGH. However, jump heights during session 1 were substantially greater than those during session 2 when using the 3 MEAN data. CVs between sessions for HIGH ranged from 4.0 to 5.6%, and those for 3 MEAN ranged from 4.2 to 5.2%. The ICCs ranged from 0.87 to 0.93 for HIGH and from 0.89 to 0.93 for 3 MEAN. Given the maximal nature of vertical jump tests, it seems appropriate to use the highest jump from a number of trials for women and men when using a contact mat. Practitioners and researchers can use the data to identify the range in which the true value of an athlete's score lies and calculate sample sizes for studies assessing height during CMJs recorded from a contact mat.