A 3-yr field survey was carried out in France, from 2002 to 2005, to study honey bee (Apis mellifera L.) colony health in relation to pesticide residues found in the colonies. This study was motivated by recent massive losses of honey bee colonies, and our objective was to examine the possible relationship between low levels of pesticide residues in apicultural matrices (honey, pollen collected by honey bees, beeswax) and colony health as measured by colony mortality and adult and brood population abundance. When all apicultural matrices were pooled together, the number of pesticide residue detected per sampling period (four sampling periods per year) and per apiary ranged from 0 to 9, with the most frequent being two (29.6%). No pesticide residues were detected during 12.7% of the sampling periods. Residues of imidacloprid and 6- chloronicotinic acid were the most frequently detected in pollen loads, honey, and honey bee matrices. Several pairs of active ingredients were present concurrently within honey bees and in pollen loads but not in beeswax and honey samples. No statistical relationship was found between colony mortality and pesticide residues. When pesticide residues from all matrices were pooled together, a mixed model analysis did not show a significant relationship between the presence of pesticide residues and the abundance of brood and adults, and no statistical relationship was found between colony mortality and pesticide residues. Thus, although certain pesticide residues were detected in apicultural matrices and occasionally with another pesticide residual, more work is needed to determine the role these residues play in affecting colony health.