The aim of this study was to evaluate the reliability and sensitivity of variables for the direct assessment of kicking performance in young soccer players. One hundred and six elite young soccer players were divided into 4 age groups (12-15 years). Absolute error (AE), variable error (VE), and constant error (CE) were evaluated as the variables of kicking accuracy, whereas the kicking velocity variables involved the maximum ball velocity (BVmax) and the ball velocity during accurate kicks (BVacc). Results suggested low-to-moderate reliability of the kicking accuracy (intraclass correlation coefficient [ICC] = 0.00-0.67) and high reliability of the kicking velocity variables (ICC = 0.87-0.94). Regarding the sensitivity, most of the variables detected the differences both between the dominant and nondominant legs and among the age groups. Because the evaluated variables should have a property of face validity, the findings obtained generally suggest that AE (and perhaps VE, as the measures of kicking accuracy) and both BVmax and BVacc (as the measures of kicking velocity) could be used within a routine composite test of kicking performance in young elite soccer players. Further development of the evaluated composite test of kicking performance could be based on the involvement of other kicking techniques and on testing the athletes of different ages, levels of skill, or sport specialization.