The main aim of this study was to examine the impact of dosing regimens on patients' persistence to bupropion. A nationally representative patient-level database comprising of pharmacy and medical claims was used to identify patients with depression (ICD-9-CM: 296.2, 296.3, 300.4, 311), who had initiated therapy with bupropion sustained release (b.i.d.; 2 doses/d) or extended release (q.d.; 1 dose/d) tablets from September 2003 to February 2004; had no previous antidepressant or benzodiazepine use; and had 9 months of follow-up. Persistence was measured using prescription claims, and calculated using the medication possession ratio [MPR; (total days supply; all filled prescriptions)/270 d]. Multivariate logistic regression compared the likelihood of achieving MPR ≥0.70 controlling for age, sex, and index date. A total of 3132 patients were included (b.i.d.: n = 2382; q.d.: n = 756). q.d. patients on average had a significantly higher MPR than b.i.d. patients [q.d. 0.52 (±0.35), b.i.d. 0.35 (±0.26)]; P < 0.001) and had a higher likelihood of achieving an MPR ≥0.70 (q.d. 35%, b.i.d. 12%, P < 0.0001). After controlling for differences in baseline characteristics, b.i.d. patients were only one-fourth as likely (odds ratio = 0.260, 95% confidence interval: 0.214-0.316) to achieve MPR >0.7. The use of the once-daily, extended release formulation of bupropion appeared to significantly improve patients' persistence to therapy for the treatment of depression.