Malnutrition and nutritional deprivation are common in hospital patients. Consequent dysfunction is exacerbated in the presence of an inflammatory state and leads to increased morbidity, mortality and treatment cost. This thesis tested the hypothesis that clinical outcome could be improved, and treatment cost reduced, if enteral nutritional (EN) was; a) initiated earlier after the pathological event, and b) when > 50% of a patient's estimated energy and nitrogen requirements were met by EN (ie. aggressive EN). A Preliminary Investigation found that patients starved for &le; 5 days compared to > 5 days, had a lower mortality (p < 0.003) and shorter duration of nasogastric (NG) feeding (p = 0.049). The population studied was heterogenous and no account was taken of disease severity. The hypothesis was therefore re-tested in burned patients, controlling for disease severity. The delay before attempting aggressive EN was associated with major complications excluding (p < 0.001) and including mortality (p = 0.018), length of (hospital) stay (LOS) (p = 0.011), and treatment cost (p < 0.001). Finally, a prospective randomised controlled trial (PRCT) was used to test the hypothesis. A study of patients undergoing major GI surgery failed due to poor recruitment, but useful aspects of this protocol were adapted for a similar trial in head-injured patients. In this group, early aggressive EN was associated with a reduction in infective complications (p = 0.0195), duration of mechanical ventilation (p = 0.019) and treatment cost (p = 0.006). In certain conditions at least, early aggressive EN is associated with improved clinical outcome and reduced treatment cost. Much of this improvement appears to be due to a reduction in infective complications.