With a sample of 29 of the best Slovenian ski jumpers, a research project was carried out with the purpose of determining the structure relation of chosen dynamic and kinematic variables during the take-off of ski jumpers. The experiment was performed in August 2008 on the jumping hill in Hinterzarten, Germany (K=95m). The subjects jumped seven times without breaks between rounds. The analysis was done on variables that determine the technique of take-off in ski jumping (in-run velocity - km/h, vertical take-off velocity - m/s, precision of take-off - cm). The criteria variable was the length of the jump (m). The variability of the long distance of the jumps was significantly strong. The reliability of all used multi-item variables was high and satisfactory in most variables (in-run velocity - 0.98, vertical take-off velocity - 0.98, precision of take-off - 0.85, length of the jump - 0.95). The factor analysis produced an independent latent structure (explanation of variance = 93.3%) of five specific factors (1. in-run velocity connected to distance jumped (39.8 % of VAR.), 2. vertical take-off velocity strongly connected to distance jumped (26.0 % of VAR.), 3. precision of take-off partly connected to distance jumped (14.9 % of VAR.), 4. precision of take-off in the 7th round (6.7 % of VAR.), 5. precision at take-off in the 4th round (5.7 % of VAR.). The present factor structure confirms the hypothetical model of three independent motor tasks to be optimally realized in the take-off of the ski jumper. Criteria variables influencing the length of jumps were mainly associated with the first two factors, which confirm the basic hypothesis that the length of the jump reflects the overall output quality of the first two factors. The accurancy factor of take-off affects the length of the jumps indirectly and latently through these two fundamental factors.