ABSTRACT: OBJECTIVE:A recent study observed a 2-fold greater risk of Parkinson disease (PD) in relation to the ?2-adrenoreceptor antagonist propranolol and a markedly lower risk of PD for the ?2-adrenoreceptor agonist salbutamol. We examined whether confounding by clinical indication for these medications, that is, tremor and smoking-related pulmonary conditions, explained these associations. METHODS:In a large, population-based case-control study of United States Medicare beneficiaries in 2009 with diagnosis codes, procedure codes, and prescription data (48,295 incident PD cases, 52,324 controls), we examined the risk of PD in relation to use of selected ? antagonists (propranolol, carvedilol, metoprolol), the ?2 agonist salbutamol, and other medications used for the same clinical indications (primidone, inhaled corticosteroids). We adjusted for demographics, smoking, and overall use of medical care. We then examined the effect of also adjusting for clinical indication and applying medication exposure lagging. RESULTS:Propranolol appeared to increase PD risk (odds ratio [OR]?=?3.62, 95% confidence interval [CI]?=?3.31-3.96). When we adjusted for tremor or abnormal involuntary movement prior to the PD diagnosis/reference date and lagged propranolol exposure, the association was 0.97 (95% CI?=?0.80-1.18). Primidone, also used for tremor, was similarly sensitive to this adjustment and lagging. ? Antagonists not indicated for tremor appeared to reduce PD risk (carvedilol: OR?=?0.77, 95% CI?=?0.73-0.81; metoprolol: OR?=?0.94, 95% CI?=?0.91-0.97) and were insensitive to adjustment for indications and lagging. Neither salbutamol nor inhaled corticosteroids were consistently associated with PD risk. INTERPRETATION:?2-adrenoreceptor agonists and antagonists do not appear to alter PD risk. Ann Neurol 2018;84:691-701.