OBJECTIVE: To examine the persistency rates and persistency days of rivastigmine and donepezil in patients with Alzheimer's Disease (AD) and to determine predictors of persistency. METHODS: This retrospective study used MarketScan® research database during 01/01/1999-12/31/2002. The study identified 229 new users of rivastigmine and 554 new users of donepezil (i.e., no use of cholinesterase inhibitor (ChEI) during the 18 month-history period) who were newly diagnosed with AD between 07/01/2000-06/30/2001, were at least 65 years old, and had continuous health/prescription insurance. Patients who refilled their initial ChEI prescription within a permissible gap of 60 days after depleting the drug supply from the prior prescription were considered to be persistent and the robustness of the persistency definition was tested using sensitivity analysis. Kaplan-Meier method and Cox proportional hazard models were performed to examine the trends of persistency and to identify factors that significantly influenced persistency. RESULTS: During the 1-year follow-up period, 53% of rivastigmine and donepezil users discontinued or switched their medications. Discontinuation was more common in donepezil-users (43%) than in rivastigmine-users (37%), but switching occurred more frequently in rivastigmine users (16%vs.10%). Rivastigmine-users continuously used their medication for an average of 234 days while donepezil-users for 235 days (p=0.91). Results were stable with variations in the definition of continuous therapy. The risk of discontinuation/switching was not statistically different between the two groups. Patients who used a CNS medication before initiation of therapy, did not visit their physician office frequently or were not hospitalized after initiation of ChEI therapy were more likely to discontinue/switch their initial ChEI. CONCLUSION: Levels of persistency among the elderly patients with AD is similar between rivastigmine and donepezil in a real-world setting. Further research is needed to determine ways of improving persistency and to estimate the long-term economic impact of persistency with ChEIs on AD treatment costs.