Attempts to protect and restore gully and bush remnants began about 20 years ago in the Hamilton Ecological District especially near Hamilton City. Research was carried out to evaluate the restoration efforts. Factors affecting the ecological failure or success of the restoration plantings were investigated. Sixty-six experimental plots were set up in both public and privately owned areas within Hamilton City and adjacent gully systems. The plots were assessed to compare vegetation change in patches planted in native species with naturally regenerating patches and mature native forest. A range of variables measured key ecosystem functional, structural and compositional attributes. Different planting and maintenance regimes and environmental factors likely to be implicated in the success of plantings were also evaluated. Analysis included ANOVA comparisons between experimental blocks and ordination and classification of the plots using principal components and dendrogram clustering. Vegetation change in the most significant variables, towards the reference mature forest ecological state was found to be more rapid in four of the experimental blocks comprised of twenty plots. Three of the experimental blocks, comprised of 12 plots, are deteriorating. In the deteriorating blocks, native species recruitment and species diversity of regeneration were low, and exotic liana species were increasing. The presence of and increase with age in the number of native lianas and epiphytes in restoration plots was generally poor. Treatments that generally appeared beneficial for patch ecological condition included good quality maintenance and low level of human disturbance. Planting a diverse range of species and enrichment planting appeared more beneficial if they were linked to good maintenance. Close proximity of seed sources was also indicated as a factor in good ecological condition of patches. Use of exotic canopy species as a nurse for restoration appeared to reduce success of the restoration. However, the canopy effect may be due to soil condition (compaction) or allelopathic effects of canopy species. In the cluster and principal components analysis, the plots clustered into three distinct groups, based on soil texture: one with heavier soils and two with lighter soils. Within each group, there was a trend towards more advanced ecological condition with age. The signature species associations identified with each cluster reflected the likely soil water availability. Species composition was shown to be related to age but less related to the functional and structural condition of restoration patches. In terms of the theoretical aspects of restoration, the findings suggest that restoration in isolated urban restoration patches, requires assistance in the form of enrichment within or peripheral plantings in the neighbourhood of the patches. The research emphasizes the importance of the human and alien species context and in particular the value of controlling human disturbance. It supports the concept of multiple restoration pathways and the concept of restoration as a means to speed up vegetation change if accompanied by appropriate management. However, the research does not support the idea of a closely defined assembly order for the vegetation community. Attending to the structure and function, particularly the regeneration and dispersal functions, of restoration patches appears to be more important for successful restoration. Composition appears to follow from appropriate structure and function. Recommendations include the following: The community should set clear goals for restoration. Species should be selected for initial establishment and future seed sources and combined with careful site-species matching. Soil conditioning should be considered on difficult sites. The choice of methods should balance rapid cover needs with the need to maintain species diversity. Maintenance should be minimally disturbing but of a high quality. Human disturbance should be actively managed. Planting within the vicinity of restoration patches, should be designed to enhance native and minimise alien species seed sources. Further exploration of the effects of exotic nurse species on restoration would be useful in order to reach conclusions that are more definite. Further research into low disturbance methods of enrichment would be valuable. Attempting restoration by enrichment of weed-dominated sites would also be an interesting trial.