Numerous bee species provide pollination services in agricultural ecosystems. Evaluating a pollinator's performance with regard to a crop is an important step in attributing pollination services and predicting how changes in a bee community or foraging environment will affect those services. We used multiple criteria to evaluate pollinators of North Carolina highbush blueberry, Vaccinium corymbosum L., agroecosystems. For five groups of bees (Apis mellifera L., Bombus spp., Habropoda laboriosa F., small native bees, and Xylocopa virginica L.), we measured forager abundance through transect observations, quantified per-visit efficiency as viable seed set resulting from a single visit, and analyzed bee presence in different weather conditions. We also considered two other criteria affecting pollinator performance-visitation rate and interspecific influence. A. mellifera was the most abundant bee in the majority of our survey sites, yet had low per-visit efficiency and reduced foraging activity in inclement weather. Small native bees were highly efficient pollinators. Their visits resulted in nearly twice as many seeds as A. mellifera or H. laboriosa. Bombus spp., H. laboriosa, and small native bees were more resilient to fluctuations in temperature, wind speed, and solar radiation than A. mellifera. Although nectar-robbing X. virginica contributed to little pollination through direct flower visits, their presence within the crop impacts the behavior and performance of other individuals. Underscoring the importance of evaluating pollinator performance via multiple criteria, our results show that bee groups contribute to pollination in different ways. These differences may provide functional complementarity and stability of pollination services to agricultural systems.