Concerns have been expressed about the reliability of clinical practice guidelines. We analyzed 3 guidelines from medical specialty societies about dronedarone hydrochloride, an antiarrhythmic drug related to amiodarone hydrochloride, for treatment of patients with atrial fibrillation. We compared the recommendations in these guidelines with the conclusions about dronedarone that we reached by applying the Grading of Recommendations Assessment, Development and Evaluation (GRADE) Method to the same evidence base. In our analysis, as a rate control drug, dronedarone was better than placebo only for a surrogate outcome (heart rate). As a rhythm control drug, dronedarone was associated with 13 (95% CI, −15 to 61) excess deaths per 1000 patients treated as compared with placebo. Compared with amiodarone, dronedarone was less effective (214 [95% CI, 130 to 294] more recurrences of atrial fibrillation per 1000 patients treated) and similarly tolerated (−28 [95% CI, −69 to 33] more serious adverse events requiring drug suspension per 1000 patients treated). Despite the limits of the evidence, all 3 guidelines recommended dronedarone for prevention of recurrences of atrial fibrillation; 2 of the guidelines recommended it as a rate control agent. Our findings raise questions about the reliability of these clinical practice guidelines, as well as the financial associations between many of the panel members and the manufacturer of dronedarone.