Restraint theory was originally developed to explain differences in eating behaviour between normal-weight and obese individuals. It represented a development from explanations based on obesity per se, and instead proposed the activity of dieting as the causal predictor of eating behaviour. Research has demonstrated that highly restrained individuals are more likely than unrestrained individuals to overeat under certain disinhibiting circumstances. The present thesis aims to investigate some of the functional effects of dietary restraint. Chapter 2 evaluates two different rationales for short-term starvation and the interpretation of the results in terms of the relative importance of the internal versus external cues suggests that external cues are very important in determining (over)eating behaviour. Chapter 3 assesses the functional role of restraint in the adolescent population and provides evidence of a restraint x disinhibitor (anxiety) interaction. Restraint is therefore functional in predicting eating behaviour even in the young adolescent population. Chapter 4 evaluates the role of imagining eating food as a potential disinhibitor and results provide an insight into the determination of highly restrained individuals when faced with a situation where it is possible to maintain high levels of restraint, and I have termed this phenomenon "superinhibition". Chapter 5 psychometrically assesses the various techniques of measuring restraint and results provide clear evidence for the use of the Restraint Scale for identifying chronic dieters. The results of this thesis are analysed in terms of current Restraint Theory, and implications for further research are discussed.