Background. This paper examines the contributions of dispositional and non-dispositional factors to post-disaster psychological morbidity. Data reported are from the 845 participants in the longitudinal component of the Quake Impact Study.

Methods. The phase 1 survey was used to construct dimensional indices of threat and disruption exposure. Subsequently, a range of dispositional characteristics were measured, including neuroticism, personal hopefulness and defence style. The main morbidity measures were the General Health Questionnaire (GHQ-12) and Impact of Event Scale (IES).

Results. Dispositional characteristics were the best predictors of psychological morbidity throughout the 2 years post-disaster, contributing substantially more to the variance in morbidity (12–39%) than did initial exposure (5–12%), but the extent of their contribution was greater for general (GHQ-12) than for post-traumatic (IES) morbidity. Among the non-dispositional factors, avoidance coping contributed equally to general and post-traumatic morbidity (pr = 0·24). Life events since the earthquake (pr = 0·18), poor social relationships (pr =−0·25) and ongoing earthquake-related disruptions (pr = 0·22) also contributed to general morbidity, while only the latter contributed significantly to post-traumatic morbidity (pr = 0·15).

Conclusions. Medium-term post-earthquake morbidity appears to be a function of multiple factors whose contributions vary depending on the type of morbidity experienced and include trait vulnerability, the nature and degree of initial exposure, avoidance coping and the nature and severity of subsequent events.