We use data from the military enlistment for a large representative sample of Swedish men to assess the importance of cognitive and noncognitive ability for labor market outcomes. The measure of noncognitive ability is based on a personal interview conducted by a psychologist. Unlike survey-based measures of noncognitive ability, this measure is a substantially stronger predictor of labor market outcomes than cognitive ability. In particular, we find strong evidence that men who fare badly in the labor market ‒ in the sense of long-term unemployment or low annual earnings ‒ lack noncognitive but not cognitive ability. We point to a technological explanation for this result. Noncognitive ability is an important determinant of productivity irrespective of occupation or ability level, though it seems to be of particular importance for workers in a managerial position. In contrast, cognitive ability is valuable only for men in qualified occupations. As a result, noncognitive ability is more important for men at the verge of being priced out of the labor market.