As a partial explanation for the maintenance of high tree diversity in wet tropical forests, Janzen (1970) and Connell (1971) independently hypothesized that natural enemies act to increase spacing within these tree populations through disproportionately high attack on progeny near adults. Both authors also hypothesized a minimum critical distance effect, because of 100% progeny mortality within a given distance of adults. We describe the necessary and sufficient conditions for testing these hypotheses, and show that attempts to evaluate them have been hampered by use of the inappropriate standard of regular spacing. Data describing the spacing dynamics of Dipteryx panamensis, a rain forest canopy tree, support both hypotheses. From 7 mo to 2 yr postgermination, seedling survival was positively correlated with distance to adult and negatively correlated with local conspecific seedling density. Partial correlation was used to separate the effects of density and distance, and it was shown that seedling density was the only significant factor in this case. Older juveniles and saplings occurred at greater distances from the nearest conspecific adult than did 1980 seedlings. No seedlings or juveniles survived within 8 m of an adult bole. A review of 24 data sets on tropical woody plants showed that most evidence indicates either density-dependence or distance-dependence in progeny mortality, as hypothesized by Janzen and Connell. Some positive evidence also exists for the minimum critical distance effect for tropical trees. In most of the cases involving seedling mortality, however, alternative causal factors such as intracohort competition or allelopathy were not ruled out. Before generalizations can be made about this process in tropical forests, carefully designed studies are needed on more populations of tropical trees.