Quantitative risk assessment affords an objective approach for assessing ecological risk from crops produced using biotechnology. Ecological risk assessment for plant-incorporated insecticidal proteins necessitates consideration of risks to nontarget insects when species-specific hazard information may be lacking. Screening-level risk assessment methods afford a means by which risks to species of concern may be evaluated conservatively using exposure estimates, host-range information, and a probabilistic estimate of toxicity to sensitive species. This approach was applied to the special case of Bt corn pollen risk to monarch butterfly, Danaus plexippus (L.), populations; the results were compared with more highly refined risk assessment techniques in terms of the risk conclusions which can be developed with more highly certain information. Exposure analysis based on readily available literature showed pollen interception by the host for monarch butterfly larvae (common milkweed, Asclepias syriaca L.) declined exponentially with distance from the pollen source. Intra- and inter-genera sensitivity of lepidopteran species was used to project effect to monarch butterfly larvae. When the 90^th percentile of effect (LC₅₀) was used to estimate monarch butterfly sensitivity to Bt corn pollen expressing Cry1A(b) protein, the risk of lethality to individual larvae was negligible at >1 m from the edge of source corn fields. Subsequent field measurements of pollen distribution, interception by milkweed, and especially effects determinations for monarch butterfly larvae exposed to Cry1A(b) toxin indicate that the screening-level approach was effective in focusing the scope of the problem to exposure from high-expressing Cry1A(b) events occurring within source cornfields or at the near-field edge. Screening level risk assessment conservatively identifies the scope of concern and the uncertainties that need clarification so that subsequent research can be appropriately focused.