OBJECTIVES: This study examined the persistence and adherence for patients using latanoprost, travoprost, and bimatoprost across multiple health plans over 12 months. METHODS: Glaucoma patients were identified from an employer-based database covering 1.8 million lives in 40 health plans. Patients with a glaucoma medical claim and a pharmacy claim for latanoprost, travoprost, or bimatoprost from September 31, 2001 through March 31, 2002 were eligible for study entry. Continuous eligibility was required 180 days prior to the index date, defined as the date of the first prescription claim for an ophthalmic drug of interest, with no evidence of ophthalmic drug use during that time. These patients were defined as "new therapy starts". Persistence at 12 months and number of days of adherence was determined for new starts with at least 3 months of therapy following the index date. Due to potential inconsistencies with days supply reporting at the pharmacy level, a clinical algorithm was developed to compute days on therapy. RESULTS: At total of 3,822 glaucoma patients were identified with at least one claim for latanoprost, travoprost, or bimatoprost. Patients were on average 73.1 years (SD = 10.1, range = 15-88) and 53.1% female. A total of 2,666 (69.8%) completed the first three months. A total of 70.1% were persistent with therapy at 12 months and were adherent 83.1% of the time. Using the quantity dispensed and the number of days between refills yielded 8 days of therapy per 1-mL of ophthalmic solution. The mean number of days on therapy for bimatoprost was significantly greater than latanoprost (p <0.05). CONCLUSIONS: This retrospective database analysis assessed persistence and adherence for glaucoma patients using latanoprost, travoprost, and bimatoprost for 12 months. Although most patients were persistent and adherent to their therapy for at least three months and then at 12 months there may still be opportunities to improve persistence and adherence with these important ophthalmic therapies.