This paper investigates the differentiation of cognitive abilities as a function of neuroticism. Specifically, we examine Eysenck and White's [Eysenck, H. J., and White, P. O. (1964). Personality and the measurement of intelligence. British Journal of Educational Psychology, 24, 197-201.] hypothesis that cognitive abilities are less differentiated at high levels of neuroticism than they are at low levels of neuroticism. In spite of its potential importance, this hypothesis has only received limited research attention. Although a recent series of papers has found some initial support for the neuroticism-differentiation hypothesis, these studies have relied on correlational techniques. The current paper uses a more rigorous structural equation modeling technique to explicitly test for measurement invariance to determine whether the structure of cognitive abilities differs between high and low neuroticism levels, as measured by two self-report scales. Results from two datasets provide convincing evidence for complete measurement invariance across different levels of neuroticism, thus disconfirming the neuroticism-differentiation hypothesis.