Forty colonies of Apis mellifera macedonica were established in Greece with naturally-mated sister queens, two frames of sealed brood and 2 kg of bees, and requeened either every year (A), every two years (B), every three years (C), or left to replace their queens through supersedure (D). All colonies were managed in the same way during the seven-year experiment. Brood area was significantly higher in groups A and B than in C or 0, except for the first two years. During the first three years there were no significant differences in honey production between any of the groups, but group 0 subsequently produced significantly less honey than all other groups. There were generally no significant differences in each year's honey production between colonies in groups A, B or C, suggesting that requeening every two or three years is adequate.