Foreign plant pathogens for biological control of weeds must be determined safe before introduction into North America. These constraints apply also to phytophagous insect candidates for biological control of weeds. The objective of risk assessments is to find candidates that attack only the target weed in North America, i.e., the candidate is host specific. There is a lack of knowledge concerning most of these North American target weeds and, in particular, their pathogens; these species are not economically or ecologically important where they are native. Furthermore, there is no information about new associations that may develop on native North American plants after introduction of a candidate agent. Demonstration of host specificity is the best evidence for the safe use of a candidate biological control agent in North America, and the host range determination is central to any such risk assessment. This involves tests on economically and ecologically important North American plant species. Acquisition of native plant species and in particular, rare or otherwise endangered and threatened species, poses a significant challenge to this effort. Seed sources for this material ranges from commercially available crop, native, and weed species, to contacts with colleagues who study the target weed or related native plants, state seed repositories (e.g., California Dept. Food & Agr. Biological Control Program), Enadangered and Threatened species recovery programs (through the U.S. Fish & Wildlife Service), Plant Introduction stations, germplasm repositories, (e.g., U.S. National Germplasm System, The Missouri Botanical Garden Center for Plant Conservation), the National Heritage Program, native plant societies and clubs (e.g., California Native Plant Society), The Nature Conservancy, or any other individual or organization that has material. Risk assessments would be facilitated greatly with better understanding of and access to native plant material.