Adolescence is often viewed as a critical period for selection in youth soccer. The present study compared the characteristics of regionally selected and non-selected under-14 players (U-14) as a group and by position. Players were classified as local (n=69) and regional (n=45). Weight, height, skinfolds, functional capacities, soccer skills and goal orientation were measured and skeletal age was assessed with the Fels method. Factorial ANOVA was used to test the effect of selection, position and respective interaction terms, while discriminant analysis was used to identify the variables that contributed to selection. Selected players had an advanced maturity status (F=24.97, p<0.01), were heavier (F=30.67, p<0.01) and taller (F=35.07, p<0.01); performed better in explosive power (F=21.25, p<0.01), repeated sprints (F=20.04, p<0.01) and ball control (F=3.69, p<0.05); and were more ego oriented (F=13.29, p<0.01). The 2 competitive groups did not differ in agility, aerobic endurance, dribbling, shooting, passing, and task orientation. Position-related variation was negligible. The percentage of players who were correctly classified in the original groups was slightly lower when the analysis was performed for the total sample (86%) than by position (86-90%). Future research on talent identification and selection should adopt a multidimensional approach including variables related to the physiological, perceptual, cognitive and tactical demands.