This study investigates the extent of pollinator sharing and specialization of three co—occurring species of milkweed (Asclepias) in experimental plantings. The potential for competition among plants for pollinators was also assessed from observed levels of pollinator constancy, interspecific pollination, and seed set in experimental populations. All three species (A. incarnata, A. syriaca, and A. verticillata) were visited by a wide variety of Hymenoptera and Lepidoptera. Niche breadth estimates for populations of A. syriaca were comparatively low (.09—.27); Bombus griseocollis and the introduced honeybee, Apis mellifera, were clearly the most common visitors to this species. Niche breadths were greater, implying less specialization, for A. incarnata (x = .30) and A. verticillata (x = .39). Paper wasps (Polistes) and digger wasps (Sphex) were numerous and effective pollinators of A. verticillata, but bumblebees comprised the largest proportion (27%) of the pollinators at some sites. Similarly, utilization of pollinators A. incarnata was highly dependent on the locality of the population studied, with Sphex pennsylvanicus, Apis mellifera, Bombus griseocollis, and Xylocopa virginica constituting the most abundant pollinators. Intraspecific comparisons of the similarity of pollinators among populations of A. syriaca gave values that were higher than corresponding values for different species within a locality, implying possible niche differentiation with respect to this species. However, a hypothesis of pollinator partitioning among species was not accepted for A. incarnata and A. verticillata because estimates of niche overlap between these species were as great as those involving populations of either species alone. Both the timing and amount of flowering, as well as the degree of interplanting of A. verticillata and A. incarnata, affected the abundance and constancy of pollinators on these species in experimental plots. Interspecific flights or pollinators and interspecific insertion of pollinia for A. incarnata were much higher in 1978 when flowering of A. verticillata was earlier and more extensive. During that same year, interspecific pollination of A. verticillata was lower. In addition, the abundance of Bombus griseocollis on this species declined rapidly as A. incarnata came into flower, even though peak flowering of A. verticillata had not been reached. Mixed plantings of these species gave higher levels of interspecific pollination and lower seed set than unispecific plantings at the same density. The lack of specialization in pollination of A. incarnata and A. verticillata is not surprising since these species typically occupy different habitats and have nectar that is readily accessible. However, the current proliferation of disturbed environments where species occur in close proximity provides the potential for exploitative competition between them when pollinators are a scarce or limiting resource. The presence of pollen from other species may also interfere with intraspecific pollination, reducing seed set. The fitness reduction may be small, however, in view of the perennial habit and abundant seed produced by plants of these species.