The New Zealand orchid flora comprises twenty-five genera and at least 100 species occurring throughout the country. Although the number of endemic species is high (69%) only four genera are endemic to New Zealand. The main physical threats to orchid survival in New Zealand are habitat destruction, modification and fragmentation. The effect of the disruption of interactions with their pollinators has never been considered. This study concentrates on this mutualistic interaction, by assessing the breeding system, pollination syndromes and pollinator-dependence of four widespread terrestrial (Gastrodia cunninghamii, Thelymitra longifolia, Pterostylis alobula and P. patens) and four widespread epiphytic orchids (Earina autumnalis, E. aestivalis, E. mucronata and Winika cunninghamii) occurring in the southern portion of the North Island. In order to determine the breeding system and the presence of self-incompatibility, hand-pollination treatments were conducted in all eight orchid species during the flowering seasons of 2001 and 2002. Pollen grains and ovules numbers, pollen:ovule ratio and presence of floral scent glands were assessed. In those nectariferous species (E. autumnalis, E. aestivalis, E. mucronata and W. cunninghamii), the nectar standing crop was determined using the anthrone colorimetric assay for total carbohydrates. The activity of pollinator was observed both in the field and in captivity. Insects observed foraging in these orchids were identified and ranked according to their likely pollination effectiveness. Finally, measurements of pollination success and pollinia removal and deposition were used to assess whether fruit-set is pollen limited in these species and explore the effect contrasting rewarding strategies (nectar v/s deception) has on the pollination success of these orchids. Pollination treatments in three terrestrial (T. longifolia, P. alobula and P. patens) and two epiphytic (E. autumnalis and E. mucronata) orchids confirmed the absence of genetic incompatibility. Despite these five orchids being self-compatible, their reproduction relies on contrasting reproductive strategies. T. longifolia is predominantly self-pollinated, whereas Pterostylis and Earina species are incapable of autonomous selfing and completely dependent on pollinators. The epiphytic species E. aestivalis and W. cunninghamii are partially self-incompatible and also completely dependent on pollinators. Agamospermy is likely to occur in G. cunninghamii but not involved in seed-production in any of the remaining seven orchids. Both terrestrial and epiphytic species showed a positive reaction to neutral red except E. autumnalis. This indicated the presence of scent glands, mainly located around the column, lip and sepal tips. Pollen:ovule ratios calculated for these species ranged from 20:1 in E. mucronata and E. aestivalis to 320: 1 in P. alobula. Of the four terrestrial orchids studied, insect visitation was observed only in P. alobula. This orchid is pollinated by male fungus gnats of the genus Zygomyia (Diptera: Mycetophilidae). Pollination by sexual deception is likely to occur in species of this genus. Numerous insects were recorded visiting the nectariferous epiphytic orchids (3 orders, 13 families). Insects considered as "probable pollinator" were Eristalis tenax (Diptera: Syrphidae) for Earina autumnalis, Dilophus nigrostigmus (Diptera: Bibionidae) for E. mucronata, and Melangyna novaezealandiae (Diptera: Syrphidae), Calliphora quadrimaculata (Diptera: Calliphoridae), the Ichneumonid wasp Aucklandella sp. (Hymenoptera: Ichneumonidae), Hylaeus sp. (Hymenoptera: Colletidae) and an unidentified weevil (Coleoptera: Curculionidae) for E. aestivalis. In W. cunninghamii the species Apis mellifera and the native syrphid flies Helophilus antipodus and M. novaezealandiae were considered as "probable pollinators". Levels of natural fruit-set were similarly low in rewarding and non-rewarding species fluctuating from 4.3% (P. alobula) to 40% (P. patens). Fruiting in these orchids is pollen limited, as supplementary hand-pollinations increased fruit set above 40% in all species except P. patens. The degree of pollen limitation varied from 0.32 (P. patens) and 0.94 (P. alobula and E. mucronata). Pollen limitation in these orchids may be caused by the simplicity of their flowers, the poor efficiency of their pollinators in depositing pollinia and the use of species-specific pollination systems (e.g. Pterostylis). The survival capability and conservation requirements of these orchids are discussed in the light of the specific reproductive requirements revealed by this study.