The objectives of this study were to investigate the influence of nectar resources on: (1) territory establishment, (2) breeding success, and (3) postbreeding dispersal in Hemignathus virens on Mauna Kea, Hawaii. I observed color-banded individuals from 1971 to 1981 and manipulated food during the 1975 breeding season by placing supplementary nectar resources in an area where breeding had not occurred in the previous 4 yr. The results of this study demonstrated that H. virens is intrinsically tied to the nectar resources available within its breeding territory and that a minimum of 2,000 μl/day (1,400 cal) of nectar must be present before territory establishment will occur. This finding was further supported by the fact that no breeding occurred in the experimental area during the 6 yr after removal of the supplementary nectar. Breeding success of Hemignathus virens was influenced by different levels of nectar availability. The most successful breeding occurred at nectar levels of 20,000-40,000 μl/day (14,000-28,000 cal). More than 35,000 μl/day (24,500 cal) of nectar within a territory generally inhibited successful breeding, presumably because too much time was spent in the defense of the nectar sources. Birds with larger amounts of nectar within their territories had concomitantly larger clutch sizes. Postbreeding dispersal, an annual phenomenon in the H. virens population on Mauna Kea, did not occur if adequate food supplies remained available during the nonbreeding period.