In studies of physical performance comprising muscle strength and power, a vertical jump is a test method that frequently is used. It is important to have access to accurate measuring tools providing data with high reproducibility. Studies have shown that body composition also may play an important part in physical performance. The purpose of this study was to determine test-retest reliability for 3 different kinds of vertical jumps and to correlate jump height with body composition. Thirty-four normally trained subjects (women n = 17) between 18 and 25 years participated. Test-retest, on 3 kinds of vertical jumps, was performed with a median of 7 days between jumps. Methods used were a countermovement jump (CMJ) on a contact mat, with and without arm swing, and an Abalakow jump (AJ) using measuring tape, with arm swing. Body composition was assessed with the use of bioelectric impedance analysis. The results showed that high intraclass correlation coefficients (ICCs) were observed between testing occasions for all 3 vertical jumps (ICC between 0.48 and 0.88). The AJ in women presented the lowest ICC. Also the correlation between CMJ and AJ was high (rs = 0.88). Moderate-to-high correlations could be shown between body composition and CMJ in women (rs = -0.57-0.76). In conclusion, very high test-retest reliability for CMJ on a contact mat was found. For the AJ using a measuring tape, ICC were overall high, but a moderate nonsignificant ICC were found in women, indicating poor reproducibility. The data from the CMJ and AJ may be compared if approximately 25% of the AJ value is subtracted. In practice, this means that vertical jump tests have high reproducibility and can be used as measures of power development.