Aims: This thesis explores the application of health psychology and behavioural economic theories to understanding adherence to medications in adult patients, to determine the most useful theoretical foundations to inform the development of adherence enhancing interventions in several countries and settings. Methods and Results: A systematic review of the literature (Chapter 2) found that components from within sociocognitive (perceived barriers, perceived susceptibility) and self-regulation frameworks (necessity beliefs and medication concerns) are significantly associated with adherence to medication; and that self-efficacy was a proximal determinant of adherence in both frameworks. A multinational cross-sectional survey of self-reported nonadherence to antihypertensive medications (Chapter 3) found that low self-efficacy and a high number of perceived barriers are the main significant determinants of non-adherence, with country explaining 11 % of the variance in non-adherence. A stated preference discrete choice (DCE) analysis (Chapter 4) found that medicine characteristics of benefit, harms and convenience have significant effect on stated persistence with medication and that psychosocial influences may modify these preferences. Concurrent application of the random utility maximisation framework and health psychology models showed that components of the theory of planned behaviour had greatest influence on probability of persistence with 5-aminosalicylic acid for ulcerative colitis. Application of intertemporal choice theory to explain nonadherence (Chapter 5) showed a weak association between time preference rates. Time preference rates, however, were associated with factors from the self-regulation framework (illness consequences and concerns). A DCE of treatment harms and benefits for treatment for epilepsy (Chapter 6) found that people with epilepsy place a higher value on reduction in harms than improvements in treatment benefit, and that patients' preferences for treatment vary by patient group. When put into the context of actual event rates this has implications the interpretation of clinical studies. An empirical study of the familiarity of conditions used to elicit time preference (Chapter 7) using propensity score matched data, found a significant familiarity with condition explained between 38-53% of the variance in time preference rates. Conclusions: Consolidation of behavioural models may provide a strengthened theoretical basis for the development and assessment of adherence enhancing interventions. A tailored approach to adherence research is required to account for country and clinical differences in preferences and behaviour.