Abstract: We performed field observations and ant exclusion experiments to determine whether Oecophylla smaragina tending homopterans on the fruit (fig) of Ficus racemosa Linn. Affects the reproduction of F. racemosa in Xishuangbanna from December 2003 to May 2004. The results of preliminary observations revealed that pollinator of F. racemosa oviposit inside the figs; five species of non-pollinating wasps attacked figs from the exterior; the number of non-pollinating wasps oviposition attempt can be quantified by counting the number of scars they made on the figs. The results of preliminary experiments showed the sticky glue barrier we used to exclude ants has no effect on oviposition of six species of wasps. When there were young figs on the sample trees, we selected two branches (a pair) which were similar in size, position and developmental phase and carried out experiments on eight randomly sampled trees. We excluded O. smaragina from one branch of each pair using sticky glue. We collected figs from trees just prior to the time that wasp progeny were due to emerge. We then counted the number of individuals of each wasp species per fig. The results showed that foraging workers of ants attacked non-pollinating wasps severely when which ovipositing on the figs. But they could not interfere with pollinators, which could enter the fig from the ostioles in a few seconds. The result of two-way ANOVA revealed that O. smaragina has significant impacts on the percentage (N%) of six wasp species breeding in the fig (n=82, F1,80>9, P<0.02). Ants present the N% of offspring of pollinator was 73.02. Ants excluded the N% of offspring of pollinator reduced to 2.14. Except for Platyneura agraensis, when ants were excluded the N% of offspring of other species of non-pollinating also increased. Ceratosolen fusciceps was sole pollinator of F. racemosa, while nonpollinating wasps competed with pollinator for galls or parasitized pollinator but play no role in pollination process. So O. smaragina can benefit F. racemosa indirectly by increasing the N% of pollinator while decreasing the number of non-pollinating wasps .