Background: the issue of diagnostic disclosure in dementia has been debated extensively in professional journals, but empirical data concerning disclosure in dementia has not previously been systematically reviewed. Objective: to review empirical data regarding diagnostic disclosure in dementia. Methods: five electronic databases were searched up to September 2003 (Medline, Embase, Cinahl, Sociological Abstracts, Web of Science). Additional references were identified through hand searches of selected journals and bibliographies of relevant articles and books. The title and abstract of each identified paper were reviewed independently by two reviewers against pre-determined inclusion criteria: original data about disclosure were presented and the paper was in English. Any disagreements were resolved by discussion until consensus was reached. Data were extracted independently by two reviewers using a structured abstraction form. Data quality were not formally assessed although each study was critically reviewed in terms of methodology, sampling criteria, response rates and appropriateness of analysis. Results: fifty-nine papers met the inclusion criteria for detailed review. Many of the studies had methodological shortcomings. The studies reported wide variability in all areas of beliefs and attitudes to diagnostic disclosure and reported practice. Studies of the impact of disclosure indicate both negative and positive consequences of diagnostic disclosure for people with dementia and their carers. Conclusions: existing evidence regarding diagnostic disclosure in dementia is both inconsistent and limited with the perspectives of people with dementia being largely neglected. This state of knowledge seems at variance with current guidance about disclosure