Seven treatments for the control of Varroa destructor (Anderson & Trueman) were tested to determine the optimum timing of miticide application. Threshold mite levels indicating miticide application were determined for three possible treatment dates: April, August, and October. The treatments were as follows: (1) fluvalinate in April, (2) fluvalinate in August, (3) fluvalinate in October, (4) fluvalinate in April and October, (5) fluvalinate applied continuously (except during honey flow) with replacement every 42 d, (6) control (no treatment), and (7) coumaphos in April. The number of miticide applications in a season had no effect on brood area or colony bee population a year after initiating the experiment. However, the absence of any treatment significantly reduced brood area and colony bee population and significantly increased colony mite population. Date of treatment had significant effects on colony mortality rates, mite levels, and brood area the following spring. When coupled with sampling and threshold recommendations, a single, late-season application of fluvalinate is as effective for the control of V. destructor as semiannual or continuous miticide applications. Treatment thresholds were recommended for ether roll and 48-h sticky board sampling methods in April (three and 24 mites, respectively) and August (14 and 46 mites, respectively) and for ether rolls in October (three mites) in cold climates.