Although worry is the central feature of Generalised Anxiety Disorder (GAD), little is known about the factors that contribute to pathological or problematic worry. In a recent cognitive model of GAD, Wells, A. (1995) proposed that negative appraisal of worrying itself (meta-worry or type 2 worry) should be distinguished from other types of worrying (type 1 worry). A central feature of this model is the idea that individuals with GAD hold rigid positive beliefs about the usefulness of worrying as a coping strategy. However, these individuals also hold negative beliefs and appraise worrying as uncontrollable and dangerous. This combination of cognitions and associated responses leads to an increased frequency and generality of worrying, and thus to the pathological worry characteristic of GAD. This paper reports a preliminary test of the hypothesis that meta-worry contributes to problematic and pathological worrying, and this relationship is independent of the frequency of other types of worry. In testing for associations between worry dimensions we controlled for overlaps with Trait anxiety, and the controllability of worrying. Results of a series of regression analyses support the hypothesis that pathological worry is associated with meta-worry and this association is independent of Trait-anxiety and type 1 worry. The clinical implications of these data are briefly discussed.