Africanized honey bees (AHBs) of Brazil and Mexico have proven to be tolerant to Varroa destructor mites. In contrast, European honey bees (EHBs: Apis mellifera carnica) at the same tropical study site are highly intolerant to these ectoparasites. A lower attractiveness of Varroa-tolerant AHB larvae has been hypothesised to be an important trait in reducing the susceptibitlity of AHBs to these mites. Thus, selection for EHB brood that is less attractive to mites is thought to be one possibility for limiting mite population growth and thus increase the tolerance of EHBs to the mite.In Ribeirão Preto, Brazil, European A. m. carnica bees and AHBs were tested with respect to their rate of brood infestation and brood attractiveness to Varroa mites. For the comparison of brood infestation rates, we introduced combs with pieces of EHB and AHB brood into honey bee colonies (18 repetitions). The relative infestation rate of EHB brood was significantly higher compared to AHB brood.The preference behaviour of single Varroa mites was tested in a laboratory bioassay where either living host stages were offered or host extracts were presented on dummies. By these tests we could confirm the preference of Varroa females for certain developmental host stages and for their corresponding extracts. In contrast to the within-colony results, Varroa mites in the laboratory bioassay showed a slight preference for AHB compared to EHB larvae.The gas chromatographic analysis revealed differences in the chemical spectrum of extracts obtained from different larvae. In accord with the results of the bioassays, we could detect stage-specific odour differences in larval cuticular compounds, including methyl esters and hydrocarbons that have been described as kairomones. None of these substances, however, revealed significant race-specific differences. Therefore, the quantity and composition of certain cuticular compounds seem to be responsible only for the recognition of a suitable host stage by Varroa females. The different infestation rates in the colonies, however, seem to be caused neither by race-specific differences in attractiveness of bee larvae nor by an extended attractive period of EHB larvae: both AHB and EHB larvae become attractive approximately 21 h before capping of the brood cell, and thus have the same window of time when they can be parasitised.Therefore differential Varroa-infestation rates are not related to larval attraction but probably are determined by other race-specific and colony-related factors.