Abstract
It is argued that personality scales of neuroticism and anxiety are tapping the same personality trait and that this is a risk factor for neurotic disorders. To see whether this trait is modifiable, a meta-analysis was carried out of therapy outcome studies which included a measure of trait anxiety or neuroticism as a dependent measure. This meta-analysis showed that all psychological therapies are to some extent effective. However, rational-emotive and related therapies produced particularly large reductions in trait anxiety/neuroticism (around 1.25 standard deviations). The possibility that rational-emotive education programmes could be used to lower trait anxiety/neuroticism as a preventive measure is discussed.