Introduction of the Andean grain chenopod (Chenopodium quinoa) into North America placed this crop within the distributional range of a related wild species,C. berlandieri. This wild species, native to the North American flora, is cross-compatible withC. Quinoa. Isozyme analysis of progeny fromC. berlandieri plants growing within and at the periphery of theC. Quinoa fields, combined with fertility assessment and phenetic comparison among putative hybrids and parental types, indicates that over 30% of progeny from wild plants growing as weeds withC. quinoa in 1987 were crop/weed hybrids. This high incidence of interspecific gene flow from crop to weed appears to be the result of asymmetric pollen flow to free-living plants from high-density cultivated populations. The observed level of crop/weed hybridization, combined with heterosis and partial fertility of F1 crop/weed hybrids, suggests that repeated annual cycles ofC. quinoa cultivation within the North American range ofC. berlandieri could produce introgressive change among sympatric wild populations. In terms of risk assessment for biotechnology, these results suggest that the breeding system may not provide an accurate indication of the potential for genetic interaction among predominately self-pollinating grain crops and their free-living relatives.