This thesis addresses biases in appraisal that are thought to contribute to the development and maintenance of social anxiety. The literature review outlines several theories of anxiety, before focussing on cognitive theories of social anxiety which predict that individuals with social anxiety have a bias in threat appraisal. Contemporary cognitive-motivational and neurocognitive theories of anxiety are then detailed and their relevance to social anxiety is described. Recent theories make efforts to integrate cognitive theory of appraisal with functional neuroanatomy, proposing that the amygdala is involved in threat appraisal. Research has provided support for the presence of appraisal biases in social anxiety and recent neuroimaging evidence suggests that threat appraisal of social cues is associated with potentiation of the amygdala in social anxiety. The utility of the startle response, in further investigating the predictions of the neurocognitive theory is described. The empirical paper investigates the predictions of neurocognitive theory that sub-cortical appraisals of social cues are associated to potentiated amygdala response. The startle response, a behavioural index of sub-cortical appraisal, was used to investigate response to social cues (neutral and fear faces) and non-social fear cues (light and dark patches) in individuals high and low in social anxiety (HSA vs. LSA). It was found that both groups had a potentiated startle response to the dark condition compared to the light and face conditions and there were no group differences in the response to social cues contrary to theoretical predictions. The findings are discussed in relation to theory and previous research findings.