The management of both diabetic and idiopathic gastroparesis often represents a substantial clinical challenge. In formulating recommendations for therapy, it should be recognized that these are based on less than optimal evidence; in particular, there are substantial deficiencies in current knowledge relating to the pathophysiology of gastroparesis, as well as the natural history of gastrointestinal symptoms, and the majority of pharmacologic trials have been short term and associated with methodologic limitations. Although the etiologic factors differ, the overall management principles are similar in the two conditions. Maintenance of adequate nutrition is pivotal, and parenteral nutrition may be required in severe cases associated with malnutrition. In patients with diabetes, rigorous attempts should be made to optimize glycemic control--hyperglycemia slows gastric emptying and may exacerbate symptoms and attenuate the effects of prokinetic drugs. Despite the relatively poor predictive value of symptoms, it is reasonable to suggest a trial of prokinetic therapy for about 4 weeks, rather than initially establishing the diagnosis by measurement of gastric emptying. However, it should be recognized that there is a substantial placebo response, a lack of evidence to support the cost effectiveness of such an approach, and that most patients will require prolonged therapy. In type 1 diabetic patients, prokinetic therapy may potentially benefit glycemic control, and this forms an additional rationale (albeit not established) for therapy. Some patients with diabetes and idiopathic gastroparesis with severe vomiting are unable to tolerate oral medication; in such cases subcutaneous metoclopramide may prove useful. Patients with intractable symptoms should be hospitalized and given intravenous erythromycin. The repertoire of prokinetic agents available in the United States is limited and includes metoclopramide, erythromycin, and cisapride (available by special program from its manufacturer); all of these drugs are associated with side effects. The use of metoclopramide may represent the first choice for chronic oral therapy, although it has been studied less comprehensively than cisapride. Combination therapy may be potentially more efficacious than the use of single agents. Dehydration and metabolic derangements should be corrected. The choice of chronic medical therapy should be individualized, taking factors such as age, presence of diabetes, concurrent medications, and comorbidities into account. In a small number of patients in whom medical treatment fails, surgery should be considered, and, if performed, done in a specialized center. A number of novel therapies, including gastric electrical stimulation, are currently being evaluated.