In an autumn colony almost-equal-to 800 newly emerged workers were checked for carrying mites, marked properly, and reintroduced into the hive. Then over a period of 30 d the marked bees were controlled every few days for mite infestation (table I). It was observed that within a short time after emergence, most Varroa females switched their carrier host bees (fig 1). Of the workers up to 3 d old almost-equal-to 15% were observed with mites, and after d 6 this infestation increased to almost-equal-to 30%. There was no difference between bees which at emergence were carrying mites or not. Since no drones were present, the female mites in choosing their carrier bee evidently considered only the age of the workers. In order to optimize their chance to invade a brood cell prior to operculation, the mites may prefer nurse bees. In this experiment there was no difference in infestation of workers beyond nursing age, probably because of the switch to winter conditions. The average mite infestation of unmarked workers sampled from brood frames was nearly the same as in the marked bees. The distribution of mites reported here for a colony is in agreement with data obtained in laboratory experiments. According to the literature, the behavior of Varroa females in selecting optimal host stages presumably depends on the pheromone bouquet of the individual bee.