<p>Eating disorders are an increasingly serious global health problem, resulting in substantial physical and mental health costs for sufferers and their families, and for society as a whole. There has been considerable research aimed at understanding and treating these conditions. Despite this, treatment for disordered eating remains relatively poor, leaving many with ongoing suffering. Although there are many reasons why this may be the case, a possible contributor not investigated by the literature is the quality of the theoretical explanations underpinning these treatments. In this thesis, therefore, I set out to examine the quality of the current dominant explanations for disordered eating upon which best-practice treatments are based. Upon evaluation, these explanations are found to be significantly lacking in quality, leading to subsequent weaknesses in their interventions. Underlying these specific problems, I identify three common meta-theoretical issues with these explanations: (1) the poor selection and description of explanatory targets, (2) the use of single explanatory models as complete explanations (i.e., the One Best Model perspective), and (3) the lack of systematic method informing their construction. In the remainder of the thesis, I propose solutions for each of these, arguing for (1) more appropriate selection and compositional explanation of explanatory foci, (2) transition away from a unified approach to explanation toward model pluralism, and (3) the development and application of a guiding methodological framework for explanation in psychopathology. Each of these solutions represents a significant shift in our approach to understanding and explaining psychopathological conditions. Their application to the specific problem area of disordered eating is likely to result in explanations for these conditions that are more valid, comprehensive, and fruitful, thereby providing superior platforms for clinical intervention and research.</p>