Black hornbills (Anthracoceros malayanus) appear to be the principal long-distance seed dispersers of Aglaia sp. (Meliaceae) at Pasoh Forest Reserve in Peninsular Malaysia. The squirrel Callosciurus prevostii removed some of the large seeds at least as far as adjacent crowns and sometimes dropped them after consuming only the orange, oily sarcotesta. It chased other squirrel species, which are probably seed predators, and hornbills out of the fruiting crown. Seeds on the ground beneath the parent crown were removed more rapidly than those farther away by rodent and (possibly) phasianid seed predators. Sitophilus sp. (Curculionidae) was also an Aglaia seed predator or parasite. Aglaia juveniles grew slowly under closed canopy, and small ones (<50 cm in height) had only 9.7 percent mortality per year during a 4-year period. Seedling survival was positively size dependent. For unknown reasons, small seedlings near the parent had a higher death rate than those at a greater distance; this was apparently a long-term pattern, because large seedlings and saplings did not occur within 10 and 35 m, respectively, of the parent tree's base. For this species an advantage of seed dispersal is avoidance of disproportionate seed and seedling mortality near the parent. Although Aglaia grows much faster in gaps than under closed canopy, it may require several episodes of growth in successive gaps before becoming reproductively mature. IN THEIR REVIEW OF SEED DISPERSAL, Howe and Smallwood (1982) remarked that little is known about its ecological and evolutionary advantages to the plant. Here we combine observations on seed dispersal and predation and juvenile mortality and spatial distribution to show how dispersal helps the tropical tree Aglaia sp. (Meliaceae) avoid seed and seedling mortality that is greater near the parent than farther away. Webb et al. (1967), Janzen (1970), and Connell (1971) suggested that such a mortality pattern, generated by seed predators, herbivores, or allelopathy, would promote the maintenance of high tree species richness in tropical forests. More recently, Connell (1978, 1979) has withdrawn his support for this hypothesis because field studies have shown that seed and seedling mortality is not invariably higher near conspecific adults. However, Clark and Clark's (1984) review of 24 data sets on mostly neotropical, woody plants showed that most evidence indicates either densityor distance-dependence in progeny mortality, as originally predicted by Janzen and Connell. Hubbell (1979, 1980) challenged the Janzen-Connell hypothesis on empirical and theoretical grounds, daiming the latter to demonstrate that spacing between conspecific adults could contribute little to the maintenance of high species richness. His theoretical analysis has been questioned by Becker et al. (1985), who consider the issue unresolved for nonequilibrium