Background: The cholinesterase inhibitors (ChEIs) rivastigmine and galantamine have been approved for the treatment of mild to moderate Alzheimer's disease in the Netherlands. Differences between ChEIs regarding persistence or the use of effective doses in daily clinical practice have been observed. However, most studies assessing ChEI discontinuation and associated determinants have been conducted in North America and there is a lack of knowledge about ChEI discontinuation and its determinants in daily clinical practice in Europe. Objectives: To assess ChEI discontinuation in daily practice in the Netherlands and to seek its determinants, including suboptimal utilization. Methods: A retrospective cohort study was performed using data from the Dutch PHARMO Record Linkage System. Included patients were aged ≥50 years at first dispensing of a ChEI, had a first dispensing of a ChEI between 1998 and 2008, had a prior medication history of 12 months and had at least one subsequent dispensing of any kind of medication. The proportion of patients who discontinued ChEIs over 3 years was determined. Cox regression was used to assess determinants for early (<6 months) discontinuation and, separately, for late discontinuation during a subsequent 30-month follow-up among those persisting with treatment for >6 months. Results: At 6 months, 30.8% of 3369 study patients had discontinued ChEIs, compared with 59.0% after 3 years. Thirty-five percent of patients taking rivastigmine reached the WHO-defined daily dose compared with 80%taking galantamine. At 6 months, compared with regular-dose rivastigmine, lowdose rivastigmine or low-dose galantamine was associated with an increased risk of early discontinuation, whereas regular-dose galantamine was associated with a decreased risk, as was concurrent use of cardiac medications, drugs for Parkinson's disease, propulsives, selective serotonin reuptake inhibitors and benzodiazepines. Associations of ChEI type/dose or comedications with discontinuation among patients persisting for >6 months differed somewhat from associations with discontinuation before 6 months. Conclusions: Fewer patients taking rivastigmine than those taking galantamine reached recommended doses. Furthermore, patients taking rivastigmine had an increased risk of early discontinuation compared with patients taking galantamine. Adverse effects leading to treatment intolerance and suboptimal utilization may have been contributing factors to these observed differences.