Self-focused information processing has become a central aspect of cognitive explanations of social anxiety disorder. Indeed, Clark and Wells (1995) theorized that the key feature of the disorder is the processing of the self as a social object. It is proposed that when socially anxious individuals enter feared situations, they shift their attention to a detailed monitoring of themselves. Such self-focused attention triggers recollections of negative experiences from the past and directs attention towards signs of threat. This interferes with the processing of the situation and leads to misinterpretation of others' behaviours. Although there is now considerable support for the Clark and Wells' model from research on imagery, post-event and anticipatory processing, there is surprisingly little evidence for memory disturbance. Unlike mood disorders, to date, deficits in accessing autobiographical memories have not been found for social anxiety. In this thesis it is argued that a possible reason for the lack of evidence of memory deficits in social anxiety is that researchers have not attempted to manipulate participants' levels of self-focus in conducting research on autobiographical memory. The focus of this thesis was on the accessibility and quality of the autobiographical memories of socially anxious and non-socially anxious individuals obtained under conditions of self-focus in comparison to responses obtained when not self-focused. The present research utilized a two-phase quasi-experimental design with a sample of 144 adults taken from an initial pool of 203 volunteers. The final sample comprised 30 men and 114 women, ranging in age from 17 to 67 years (M = 26, SD = 11.91, median age = 20 years). The sample consisted of 292 undergraduates and community participants recruited via advertisements. Group membership was based on selection criteria from a measure of social anxiety, the Social Phobia and Anxiety Inventory (SPAI; Turner, Beidel, Dancu & Stanley, 1989b), depression and general anxiety from the Depression, Anxiety and Stress Scale (DASS; Lovibond & Lovibond, 1995), and the anxiety disorder interview schedule (ADIS-IV-L; Di Nardo, Brown & Barlow, 1994). Initial respondents were allocated to socially anxious (SA), socially anxious depressed (SAD), and non-socially anxious (NSA) groups, with the SAD group forming a further control group of respondents with high scores in SPAI social anxiety and DASS depression. Approximately half of each group was primed to Self-focus, by making an impromptu speech which was videotaped and replayed to each participant individually, following an adaptation of a priming methodology by Perowne and Mansell (2002). The other half of the participants was not self-focused. Rather, they viewed video tape of a confederate giving a speech so as to Other-focus. The priming was prior to participants' provision of written responses to positive and negative (social anxiety) cue words using Williams' (2002) Autobiographical Memory Test. The quality of the memories was measured according to overgenerality for positive and negative cues. Reaction time was recorded for memories to the same valenced cues. The incidence of anxiety and depression content in the memories was determined using coding schemes devised by Gottschalk and Gleser (1969). As hypothesised, the results confirmed that greater levels of generality and longer reaction time to positive cues (not for negative cues) were evident when comparing self-focused SA with self-focused non-socially anxious (NSA) individuals, reflecting findings for other emotional disorders such as depression. Self-focused socially anxious (SA) individuals responded with greater levels of generality and longer reaction times for memories for valenced cues, particularly positive ones, than other-focused SA individuals. Further analysis for generality found that self-focused SA respondents were more general than other-focused socially anxious with comorbid depression (SAD) individuals for memories for positive cues. However, both groups were similar for negative memory cues. As confirmation for memory deficits as a function of the priming manipulation, greater generality and longer reaction times for memories for positive cues occurred when comparing self-focused SA individuals, to all control groups (NSA groups and other-focused SAD individuals). Without priming, SA individuals did not demonstrate autobiographical memory deficits in quality or accessibility compared to NSA controls. Exploration of the autobiographical memory content indicated that SA (compared to SAD and NSA) individuals, irrespective of focus, reported more depression than anxiety content in their memories. Additionally, self-focused SA in contrast to other-focused SA individuals, experienced heightened depression content in their memories, although none of the individual categories of depression was prominent. Predictions of SPAI social anxiety and DASS depression scores from the significant predictors associated with the memories, were undertaken separately for self- and other-focused individuals. Reaction time to positive cues was found to be the only stable predictor of both social anxiety and depression for the self-focused respondents. For the other-focused condition, only total anxiety content theme was found to predict social anxiety, albeit poorly. Overall, the present thesis established that the self-focusing manipulation unearthed a memory processing deficit in socially anxious individuals, similar to that found in most emotional disorders in terms of quality, accessibility, and content, of the memories. This result was particularly apparent in relation to responses to positive cues. The findings are consistent with the Clark and Wells' (1995) model of the disorder emphasising the role of the views of the self as the core element of the disorder. The results also show that the processing deficits of social anxious individuals are confined to the self-focus condition. It is suggested that more attention should be placed on the self-role in social anxiety and that mindfulness-based cognitive therapy, efficacious in mood disorders by reducing the overgenerality effects, is also relevant to social anxiety disorder. Future longitudinal research should be conducted with clinical groups of socially anxious patients, using a self-focus manipulation, to ascertain if the results of the present can be replicated and extended. Pre- and post-treatment measurements of the autobiographical memories of socially anxious patients should be measured and compared, to ascertain whether the deficits can be repaired through the use of mindfulness-based cognitive therapy (MBCT; Williams, Teasdale, Segal & Soulsby, 2000). Self-focus priming could also be used with other anxiety disordered individuals such as generalised anxiety disorder and panic individuals, as they too, appear to be likely candidates for the ruminative self-focus which has contributed to the memory processing deficits established in the autobiographical memories in social anxiety disorder.