Background: Generic alendronate was approved in the United States on February 6, 2008. Medicare beneficiaries might pay for generic alendronate out-of-pocket without having claims submitted, resulting in misclassification of generic alendronate use in Medicare data. Objectives: To estimate the completeness of generic alendronate use in 2008 Medicare Part D data; to identify factors associated with staying on branded alendronate versus switching to a generic product. Methods: We identified Medicare beneficiaries highly adherent (medication possession ratio ≥80%) with branded alendronate during 1/1/06-2/6/07 ("2007 cohort") and during 1/1/07-2/6/08 ("2008 cohort"). The outcome was medication status at the end of follow-up (12/31/2007 or 12/31/2008), classified as continued branded alendronate, switched to generic alendronate, switched to another bisphosphonate or presumed discontinued bisphosphonate therapy. Cox regression estimated the hazard ratio (HR) for discontinuation in 2008 compared to 2007. Multinomial logistic regression identified factors associated with medication status for the 2008 cohort. Results: Among 15310 subjects using branded alendronate in the 2008 cohort, 81% switched to generic alendronate. The proportion presumably discontinuing bisphosphonate therapy was 8.9% in 2008 compared to 7.7% in the 2007 cohort (adjusted HR, 1.15; 95% confidence interval, 1.05, 1.26). Factors associated with staying on branded alendronate in 2008 were higher income, eligibility for a low income subsidy and use of Fosamax® plus vitamin D. Conclusion: Evaluation of Medicare prescription drug data suggests that the amount of missing claims for generic alendronate in 2008 was not substantial, and misclassification of exposure in studies examining alendronate use post-generic product availability should be minimal. © 2012 John Wiley & Sons, Ltd.