The foliage-as-fruit hypothesis argues that some plants may actually promote herbivory by large ungulates at the time of seed set to achieve seed dispersal. I examined a minimum condition of this hypothesis, that large herbivores actually ingest and pass small seeds from non-fleshy-fruited plants, by screening the dung of large indigenous mammals at the time of seed set. In a lowland reserve in South Asia, I examined the dung of Greater One-Horned Rhinoceros, Indian bison, and four native cervids for seeds from plants not producing fleshy fruits but found no obvious examples supporting the foliage-as-fruit hypothesis. Frequent, predictable, and widespread disturbance of grasslands by annual flooding and periodic fires provides sufficient microsites for the propagation of annual plants. Under this disturbance regime, selection pressure for megafauna-mediated dispersal, as described by the foliage-as-fruit hypothesis, is not apparent in plants of South Asian flood plains. JANZEN (1984) SUGGESTED THAT THE FOLIAGE of smallseeded plants may function ecologically as fruit, attracting large herbivores just as arils and fleshy berries attract and reward frugivores. In this intriguing evolutionary scenario, large ungulates, lured by a nutritious fodder bait, ingested minute seeds and fruiting stalks cloaked in leaf tissues. Long coexistence with a herbivorous megafauna provided small-seeded plants with a unique dispersal guild of wide-roaming mammals that defecated seeds into favorable germination sites (e.g., edges of game trails, small forest openings, wallow edges, river and stream banks). Implicit in the foliage-as-fruit (FAF) hypothesis is the prediction that plants subjected to herbivory are compensated by effective dispersal and recruitment away from the parent. Selection for tough, digestion-resistant seeds would increase survival past the Scylla and Charybdis of megafaunal dispersal: massive, grinding molars and lengthy gut