Mānuka honey has been identified as having non-peroxide antimicrobial activity (NPA). This medical benefit has led to mānuka honey becoming a major export for New Zealand. Hives are currently located in preserved or regenerating mānuka bush. With the growth in the mānuka honey industry, interest has arisen to determine why mānuka trees produce the non-peroxide active honey. Identification of this reason will allow mānuka plantations to be planted that are expected to provide honey with a high level of non-peroxide activity. These plantations will also allow marginal land to become more productive due to mānuka being a resilient plant able to grow under harsh conditions. It has been shown that the majority of the non-peroxide activity arises from the presence of methylglyoxal (MGO). In fresh honey a minimal amount of MGO was present, but a large amount of dihydroxyacetone (DHA) was found. Previous work has shown that DHA can undergo chemical conversion into MGO under the correct conditions. A similar conversion is postulated to take place in honey, testing suggests that this conversion is a non-enzymatic reaction. Preliminary surveys have been carried out which have identified DHA in the nectar of mānuka owers. This thesis describes a wider survey of mānuka trees around New Zealand.The trees were sampled in the flowering seasons of 2009 and 2010 between October and January. Flowers were picked and frozen for processing and an aqueous soaking method was developed to extract the DHA and sugar from a pooling of 20 owers. Analyses of the samples were carried out by gas chromatography with flame ionisation detection. This method was further improved to include the nectar extraction and measurement of DHA levels within a single flower. To allow the DHA to be related back to honey, it was measured in ratio to the total sugar (Tsugar) in the nectar to give the ratio DHA/Tsugar. It was confirmed that DHA/Tsugar measured in the nectar of the mānuka flower does vary within and between the regions surveyed. Suggested causes of within region variation are the age of the tree, micro-environments and possibly genetics. Variation between regions is strongly suggested to be genetically linked. Using the work by Adam set al.(2008, 2009), it was possible to predict honey NPA values based upon the DHA/Tsugar found in the nectar and these values were comparable with the measured NPA of the honey as supplied by beekeepers. Only a poor correlation of DHA/Tsugar was found with the soil components measured by Kiefer(2010); with the leaf oil components measured by Janusch(2010) some correlation was found, when these were correlated across all the sampled regions. When each region was correlated individually, the correlation proved much stronger, suggesting a link, though most likely indirect, to the mānuka oil chemotype. Using these survey results, mānuka trees have been identified for the purpose of breeding and on-going study.