Data from the Longitudinal Aging Study Amsterdam were used to study the relationship between neuroticism and aging. At baseline, cross-sectional analyses of data from 2,117 respondents (aged 55-85 years, M = 70) showed no significant age differences. The magnitude of the 3- and 6-year stability coefficients was high, and 12% of the elderly participants showed a clinically relevant mean level change. Longitudinal multilevel analyses showed a small but statistical significant change with aging, but the mean change was not considered clinically relevant. A U-formed course was found, showing a slight decrease until respondents reached the age of 70. Adjusting the model for physical health-related variables slightly increased the stability. An additional interaction analysis showed that the individual trajectory of neuroticism was not affected by the physical health status. In conclusion, neuroticism remains rather stable in middle and older adulthood, with some apparent increase in late life.