The efficacy of drone brood removal for the management of Varroa destructor Anderson & Trueman in colonies of the honey bee, A. mellifera L., was evaluated. Colonies were treated with CheckMite in the fall of 2002. The following spring, quantities of bees and brood were equalized, but colonies were not retreated. The brood nest of each colony consisted of 18 full-depth worker combs and two full-depth drone combs. Each worker comb had <12.9 cm² of drone cells. Standard management practices were used throughout the season. Colonies were randomly assigned to one of two groups. In the control group, drone combs remained in place throughout the season. In the treatment group, drone combs were removed on 16 June, 16 July, 16 August, and 16 September and replaced with empty drone combs (16 June) or with drone combs removed on the previous replacement date. In the early fall, the average mite-to-bee ratio in the control group was significantly greater than the corresponding ratio in the treatment group. Drone brood removal did not adversely affect colony health as measured by the size of the worker population or by honey production. Fall worker populations were similar in the two groups. Honey production in treatment colonies was greater than or similar to production in control colonies. These data demonstrate that drone brood removal can serve as a valuable component in an integrated pest management program for V. destructor and may reduce the need for other treatments on a colony-by-colony basis.