The objectives of this study were: (i) to analyse anthropometric parameters, physical fitness, and throwing velocity of handball male elite youth players of different ages; and (ii) to develop a multivariate model that explains throwing velocity. Fifty-three handball men players (17.99±1.68 years old), members of the Icelandic National Teams, participated in the study. The participants were classified into the U21 National Team (n=12), U19 National Team (n=17), and U17 National Team (n=24). All were evaluated by basic anthropometry (body height, body mass, body mass index), physical fitness tests (counter movement jump, medicine ball throw, hand dynamometry, 10 m and 30 m sprint, yo-yo IR2 test) and ball speed after various handball throws at goal (a 7-m throw, a 9-m ground shot after a three-step run-up, and a 9-m jump shot after a three-step approach). A one-way analysis of variance with a Bonferroni post - hoc test was used to establish the differences between the teams. Multiple linear regression was used to predict the speed of the ball from each of the three shots taken for each team. There were no differences between the U21 and U19 teams except for the medicine ball throw, but the U19 team scored better than the U17 team in almost all variables. Ball speed after a handball shot was predicted (between 22% and 70% of accuracy) with only one or two physical fitness variables in each model ‒ medicine ball throw (in four models), counter movement jump (in two models), and 10 m sprint (in two models), being the variables that were most selective.