We observed birds and mammals visiting individual plants in a Panamanian population of the rainforest tree Virola surinamensis (Wild Nutmeg) to determine (1) relative importance of different dispersal agents, and (2) the relationships between crop size, crop quality, and seed removal. Arils of this tree are exceptionally nutritious (63% fat, 9% nonstructural carbohydrate, 3% protein), but draw remarkably few species of fruit—eating animals. Six birds take 77% of the seeds handled by animals. One toucan (Ramphastos°swainsonii) accounts for 35% of the seeds handled; other birds (Penelope purpurascens, Trogon massena, Baryphthengus martii, Pteroglossus torquatus, Ramphastos sulfuratus) play lesser roles. The smallest visitor, a cotinga (Tityra semifasciata) eats arils but does not disperse seeds. One monkey (Ateles geoffroyi) removes 3% of the seeds handled by frugivores, but smells and rejects or otherwise wastes 9%. Such a small assemblage is anomalous in a forest with at least 78 species of fruit—eating animals; other trees with arillate seeds of similar size draw as many as 23 frugivores. We hypothesize that the size of the assemblage is limited by seed size, small crop size, and perhaps by defensive compounds in the arils. Despite the small assemblage, an average of 62% (range: 13—91%) of the fruits available to animals are taken. Fruits are so nutritious that trees with both small and large crops are depleted; no variance in the percentage of seeds removed can be attributed to crop size. Fruits of different individual trees do vary in size; 59% of the variation in the percentage of seeds taken can be attributed to mean seed size. Fruit—eating animals preferentially deplete trees with small seeds, and coincidentally favor trees with high aril—to—seed ratios. No significant portion of the variance in dispersal success can be explained by aril mass alone. We hypothesize that consistent differences in seed size could result in dramatic differences in dispersal, and hence reproduction, over the lifetime of individual trees. On the other hand, selection may alternatively favor small seed size and high dispersability in years when frugivores are abundant, and large seed size and enhanced seedling vigor in the parental stand when fruit—eating birds are scarce. Small—seeded plants are likely to colonize new sites; large—seeded individuals are likely to produce offspring that fare well in competition with other seedlings near parent trees.