The main purpose of the present work was to study neurocognitive performance of adolescents at risk for emotional difficulties. The sample included a total of 1,509 adolescents from stratified random cluster sampling. Derived from this sample, a group of high-risk (n = 92) and a comparison group (n = 92) were selected based on the short version of the Positive and Negative Affect Schedule (PANAS) for comparison on the University of Pennsylvania computerized neuropsychological test battery for children (PENN). A Multivariate analysis of covariance (MANCOVA) was performed taking the scores on the PENN as dependent variables and the two groups derived from the scores of the PANAS (at risk vs. comparison) as a fixed factor. Adolescents at high risk of presenting affectivity problems showed statistically significant differences in several different neurocognitive domains, in accuracy, λ = .820, F
(9, 160,000) = 3.913, p < .01, partial η² = .180; speed, λ = .502, F (5, 88,000) = 17.493, p < .01, partial η² = .498; and efficiency, λ = .485, F
(4, 89,000) = 23.599, p <.01, partial η² = .515. The high risk group showed lower neurocognitive performance than the comparison group. In addition, a positive statistically significant correlation was found between all the neurocognitive competences (p < .05). Results found in this study reveal that neurocognitive impairments can be shown in adolescents at psychometric high risk for emotional problems before transition to more severe psychological problems.