This paper describes the construction, refinement and implementation of a self-administered measure of personal hopefulness, the Hunter Opinions and Personal Expectations Scale (HOPES). Initial state and trait versions of the HOPES instrument were utilized in three separate studies, comprising a medical student sample (N = 211), an adolescent male sample (N = 280) and a psychiatric hospital staff sample (N = 318). A revised 20-item, two factor, trait version of the scale was then utilized in a prospective, longitudinal investigation (N = 753) of the psychosocial sequelae of the earthquake which struck Newcastle (Australia) in December, 1989. Data from all four studies provide strong support for the HOPES instrument's construct, concurrent and predictive validity. Global personal hopefulness (GPH) was shown to be an enduring characteristic of individuals, with a test-retest correlation of r = + 0.71 (over 64 weeks). The association between GPH and trait anxiety (r = -0.64) raised the possibility of redefining anxiety as hope under threat. The hope subscale (HS) and the despair subscale (DS) were moderately negatively correlated (r = -0.32), suggesting that hope and despair are not simply polar opposites. There were no gender differences in GPH scores, however, there were relatively clear age effects, with those aged 70 years and over reporting the lowest levels of personal hopefulness. GPH was negatively correlated with post-earthquake scores on the General Health Questionnaire (r = -0.33), the Impact of Event Scale (r = -0.33), the Beck Depression Inventory (r = -0.54) and the global symptom index from the SCL-90-R (r = -0.43). Overall, the contribution made by personal hopefulness to post-earthquake morbidity was equal to the contributions made by initial exposure to disruption and threat experiences.