Low capsule set is a major factor limiting seed production in Eucalyptus globulus seed orchards. Trials were conducted in E. globulus seed orchards in Tasmania to identify the timing of capsule development and abortion, as well as the effect of pollination type, female tree, site, fertilisation, flower physical attributes, flower density, irrigation and weather events on capsule set and overall reproductive success. Controlled pollinations (CP), mass supplementary pollinations (MSP), open pollinations (OP) and un-pollinated controls (UP) were carried out at two orchard sites (Cambridge and Ridgley). Ramets of the same genotype differing in flower density and irrigation treatment were also compared for capsule set under MSP in three other experiments. No capsules were set in the UP, and capsule set was significantly lower following CP than OP and MSP. The genotypes at the lower altitude Cambridge site had a higher level of capsule set than the same genotypes pollinated at the Ridgley site. The major period of capsule abortion occurred between 20 and 80 days after pollination, coinciding with the period of capsule growth. There was a positive correlation between the number of fertilised ovules per aborted capsule and the length of time capsules were held on the tree. Flower physical attributes were found to correlate with reproductive success. Increased flower density had a negative impact on capsule set, as did irrigation. There was no clear evidence that weather events affected the rate of capsule abortion over the study period. Given that capsule abortion occurred during a period of fruit growth, that capsules with the lowest number of fertilised ovules aborted first and there is an impact from flower density and irrigation level, it is argued that fertilisation level and the level of resource competition are major factors determining capsule abortion.