The effectiveness of donepezil in routine clinical practice remains uncertain, and the external validity of the clinical trials to date is, necessarily, limited. It is not clear how response to treatment should be measured in routine practice, although dose escalation and continuation of treatment will demand accurate assessment. It may be argued that the size of benefits is so small that the risk of greater side effects in more typical people with SDAT, many of whom have co-morbidity and may already be on other drugs, make it difficult to conclude on overall effectiveness. Evidence is limited in quantity and by restriction to select population. The author concludes that there is evidence of temporary benefit but at considerable cost. Further research is needed in this area.