The relationship between trait neuroticism, life events and depressive symptoms was studied in a sample of New Zealand women over a 4-year period. These results were analysed using a structural equation modelling approach which attempted to estimate the effects of both random and systematic errors of measurement on the correlation between neuroticism and depressive symptoms. The analysis suggested the following major conclusions. — Firstly, simple correlations between neuroticism and depression measures are strongly biassed upward as a result of systematic errors of measurement. The correlation between neuroticism and depression was + 0.50 after correction of systematic measurement error, whereas prior to this correction it was + 0.70. — Secondly, there were consistent correlations between neuroticism measures, corrected for random and systematic measurement error, and life events scores, indicating that subjects who scored high on neuroticism had a greater susceptibility to adversity. — The implications of the approach for understanding the relationships between measures of personality traits and psychopathology are discussed.