This thesis investigates the relationship between emotional intelligence (EI), leadership styles and outcomes of leadership, within the workplace environment of the National Health Service (NHS) in the UK; which is endemic with change (Allen, 2009; Iles & Cranfield, 2004). The broad question posited is: Within the context of change, what is the nature of the association between EI, leadership styles and leadership outcomes, from the perspective of both leaders and their followers? There is a dearth of published research on EI and leadership within the context of change which makes this study particularly timely. The study employs two different models of EI that have evolved from ‘ability-based’ (Palmer & Stough, 2001; Mayer & Salovey, 1997) and ‘personality-based’ (Bar-On, 1997; Higgs & Dulewicz, 2002) theoretical perspectives. This is the first substantive study to have embraced both models and the intention here was to identify similarities and differences between the two perspectives in the context of organisational change leadership. Although there is extensive research on leadership, there still exists an acute need for EI and leadership research based on follower perspectives (Lindebaum & Cartwright, 2010; Notgrass, 2010). This study addresses this gap by investigating both leader self-perceptions and follower-perceptions of their leaders. This study has been conducted in two phases. Phase 1 focuses on leader self-perceptions. Phase 2 focuses on the combination of leader self-perceptions and follower-perceptions of their leaders. In Phase 1, this thesis postulates that within dynamic environments such as the NHS, different types of linkages can be argued between EI and the different leadership styles (Transformational Leadership, Transactional Leadership and Laissez Leadership) while suggesting that EI will predict Transformational Leadership. Furthermore, this thesis postulates that EI and Transformational Leadership will have a positive impact on leadership outcomes. In Phase 2, this thesis considers both leader and follower perspectives and identifies leaders who overestimate, underestimate or are in-agreement with followers, in evaluating their own Transformational Leadership capabilities. Based on this, the leaders are classified into self-other-agreement categories of overestimators, underestimators, in-agreement/good and in-agreement/poor leaders. Thereafter, this thesis postulates varying levels of leader EI and Outcomes of Leadership (depending upon their self-other-agreement categories) as adjudged by their followers. Similar differences across the self-other-agreement categories of the leaders are predicted in terms of the linkage between EI and Transformational Leadership as perceived by the followers. The ontology of this research is realist and the epistemology is positivist (Burrell & Morgan, 1979). Data has been collected in two phases - from leaders in the NHS and then their reporting staff. Phase 1 target population was identified as NHS staff in leadership positions involved in implementing change management initiatives. Phase 2, target population was identified as the direct reports of phase 1 participants. Phase 1 entailed purposive judgment sampling followed by random sampling. Phase 2 adopted convenience sampling. Both phases were completed through surveys. In phase 1, leaders completed a self-report of the Multifactor Leadership Questionnaire, the Swinburne University Emotional Intelligence Test (SUEIT) and the Higgs & Dulewicz Emotional Intelligence Questionnaire (EIQ). The Marlowe-Crowne Social Desirability Scale has been employed to check for possible response bias. Self developed research items were employed to obtain information regarding the extent and nature of the respondents’ involvement with change. Leaders, who consented to participate in phase 2, identified 3 to 5 reporting staff for participation. In phase 2, reporting staff completed rater-forms of the Multifactor Leadership Questionnaire and the EI 360 degree questionnaires. A number of qualitative interviews have been conducted with leaders to obtain contextual data regarding the nature of the changes, impacting upon the lives of NHS staff. Phase 1 findings based on self-ratings of leaders indicated a strong positive relationship between EI and Transformational Leadership, and between EI and one factor of Transactional Leadership (contingent reward). Contrary to the hypothesis, a negative relationship was found between EI and Laissez Faire Leadership. Regression analysis revealed that both the EI models significantly predicted Transformational Leadership. Furthermore, both EI and Transformational Leadership demonstrated a strong positive relation with Outcomes of Leadership. This study also reports higher Transformational Leadership scores for females and higher EI scores for females according to one of the Swinburne University EI model. Supplementary findings showed that; while predicting Transformational Leadership, the Higgs and Dulewicz EI model had incremental predictive validity over the Swinburne University EI model. Phase 2 findings based on follower perceptions confirmed the postulation that overestimators demonstrate lower EI and Outcomes of Leadership than other leader categories. Underestimators were perceived as manifesting higher EI and higher Outcomes of Leadership. The link between leader self-ratings of EI and follower-ratings of their leaders’ Transformational Leadership across the self-other-agreement categories was also examined. Findings supported arguments that EI and Transformational Leadership will be significantly associated for overestimators but not for underestimators. Some contradictory results were generated for the in-agreement/good/poor categories in terms of EI and Transformational Leadership correlations. These contradictory findings may be attributable to the difference in the factorial constitution and psychometric properties of the two EI models. More research is also recommended on the in-agreement sub-categorisation of focal leaders as good/poor. This study provides empirical evidence suggesting that EI predicts Transformational Leadership, effective in the context of change. EI also positively correlated with aspects of Transactional Leadership, helpful in successfully leading change. Therefore, investing in training and development of leaders’ EI has the strong potential improve the ability of change leaders. Underestimators were rated by followers as the highest on EI and Outcomes of Leadership, while overestimators were adjudged as the lowest on EI and Outcomes of Leadership. Therefore, underestimating leaders are arguably most suited to effectively lead dynamic change, as in the NHS. Female leaders manifested higher Transformational Leadership and also higher EI on one of the EI models. This provides some justification to enhance leadership responsibilities of women within transformational environments like the NHS. The clear indication that high EI predicts Transformational Leadership and high EI relates to lower Laissez-Faire Leadership can have recruitment and selection implications favouring the appointment of individuals with high EI in the NHS facing endemic change. However, there is a need to exercise caution and not simply use EI measures alone for placements and recruitment/selection. Nonetheless these results and the measures of EI and leadership could be beneficial in self development, career counselling and other social contexts in the National Health Services.