A group of 55 polish javelin throwers in junior category (18-years-old) took part in this research. For the purpose of the verification of usefulness of Kohonen’s Neural Models, statistical analyses were made of the measurement results, of the javelin throwers from the full take off. In predicting the value of variables measured with so called strong scale regression models, numerous variables were used. The construction of such models required strict determination of endogenous variable (Y- distance of javelin throw) as well as the right choice of variables in explaining the study's phenomenon. The optimum choice of explanatory variables for the Kohonen’s networks was made on the grounds of the regression analysis. During statistical analysis of gathered material neural networks were used: Kohonen's maps (data mining analysis). The obtained model has the form of a topological map, where certain areas can be separated, and the map constructed in this way can be used in the assessment of candidates for sport training.