Seed dispersal is a critical phase for individual trees because it allow seeds dispersed away from density-dependent sources of mortality, for populations by contributing significantly to neighborhood size, and for metapopulations (sensu latu) by allowing colonization of new sites. The challenge, however, is that the process that we are most interested in is the tail and that is the outcome that is most difficult to track. Parentage analysis has made great strides lately for studies using the maternal tissue in the seed or fruit to genotype seed source. However, that method has the constraint of needing mapped locations for all possible seed sources. Here, we demonstrate that the genetic signature resulting from dispersal can provide valuable information on the pattern of dispersal. Using dispersed seedlings of California valley oak found in a natural savanna setting, we address these questions: (1) What is scale of dispersal for dispersed seeds? (2) Is there evidence of long-distance seed dispersal within the study site? (3) What is the pattern of spatial autocorrelation in dispersed seeds versus seedlings and what does this indicate about the neighborhood size after the two life history stages? To address these questions, we will use data from seven microsatellite nuclear genetic markers for approximately 250 individuals. Our results indicate that dispersal is extremely restricted in this species, which would result in small effective neighborhood sizes. At the same time, we observed occasional long-distance event, which could be helpful for maintaining connectivity in a metapopulation.