Traditionally the determination of the giftedness has performed with traditional intelligence tests. The main argument to do so has been the temporal stability of this measure. In recent years some authors defend a determination early giftedness in children, although others point out that toddlers evaluation produces a significant number of false positives due to the variability in the measurement of intelligence by different factors such as plasticity, stimulation, etc.., therefore proposed the use of complementary indices for giftedness diagnosis. This paper presents a longitudinal study of two years to 49 children aged between 5 and 9 years, initially identified as gifted, to check the stability of your IQ and other measures such as the learning potential and working memory. The results show how learning potential and working memory measures remain stable in time while the IQ of a group of younger children is not maintained during this period. These results indicate the usefulness of P.A. measures as a supplementary index in the determination of the giftedness on young children.