Deer farming in New Zealand has been developed for about 25 years, with about 1.3 million deer being currently farmed. Knowledge of deer health and production has been acquired from experimentation in research stations and from field experience, but until now there has been limited research data from commercial farms. Little was known about the relative importance of a range of management practices, farm or animal characteristics, and how they related to deer health and production. "Herd health and production profiling" is a tool for studying complex systems such as pastoral farming. This is the first time this technique has been used for the study of deer farming. It is a longitudinal observational study with an holistic epidemiological approach. It involves gathering a variety of relevant information about the whole farm which is analysed by advanced multivariable statistical techniques at farm, herd or mob, or individual animal levels. The present study aimed to explore the health and production results from selected red deer farms; to investigate risk factors associated with the most important performance outcomes, and to identify the most relevant fields for further research. This study involved observations from commercial red deer farms in the lower part of the North island of New Zealand, and was conducted over two years beginning in March 1992. Potential participating farmers were solicited by mail questionnaire. Fifteen with the appropriate stock and facilities were selected after farm visits. Farm characteristics were recorded by direct measurements or by questionnaires. Farm layout was mapped. Each deer paddock was described with characteristics such as area, topography, water supply, type of fence, shelter, exposure to wind, pasture type and weeds recorded. Deer were identified by sex (hinds or stags) and age (weaners, yearlings or adults). About 2700 hinds, 2400 weaner deer and 1500 stags were individually monitored each year. Farmers recorded daily farm management practices such as grazing management, mating and calving management, handling and disease control; individual deer data such as rearing performance of hinds, bodyweight, velvet antler grade and weight; health problems and weather data. Post-mortem investigations were conducted by researchers, local veterinarians or farmers themselves. Farm visits were scheduled in March, June, September and November to record information and collect samples with additional visits as required. Body measurements were recorded, and blood and faeces samples were collected from 30 selected deer of both sex and three age categories per farm. Pasture and soil samples were collected and measurements were taken for each paddock. A body condition scoring (BCS) system was devised, and hinds scored in March before mating, September before calving and in March of the next year at weaning. All hinds mated were pregnancy tested by rectal ultrasound during the June visit and pregnant hinds were classified as having conceived before or after May 1. Blood on EDTA was analysed for standard haematology. Sera were analysed for total protein, albumin, phosphorus, gamma glutamyl transferase, blood urea nitrogen, glutathione peroxidase, pepsinogen, copper, vitamin B12, and serological evidence of yersiniosis and leptospirosis. Faeces were analysed for egg and larval counts. Pasture mineral profiles (Cu, Co, Se, Mn, Mo, Zn, Fe, and S) and soil fertility (Olsen P, K, S and pH) were investigated. Data were entered into a relational database. Databases were corrected for erroneous and missing information by reference back to farmers, and calculations were performed for mob and farm characteristics. Data on farm and deer characteristics, management practices, deer health and productivity parameters, and biological markers were tabulated and graphed at the individual and farm level. Some of the key health and production outcome were: 84.7% and 96.8% of yearling and adult hinds conceived at mating, respectively, and 17.0 and 9.1% of them lost their progeny by weaning; mean bodyweights of weaner standardised on April 1within farms ranged 42-59 kg and 39-51 kg in stags and hinds, respectively; the percentage of stags reaching 92 kg bodyweight by one year of age within farms ranged 0-48%; the mean yearling hinds bodyweight at mating was 81.3 (SD=8.5) and 82.8 kg (SD=9.4) in 1992 and 1993, respectively; 2-year-old stags produced on average 1.16 and 1.19 kg of velvet antler in 1992 and 1993, respectively; annual mortality rates of weaners (3-15 months), hinds and stags were 5.87, 1.77 and 2.60 per 100 deer-years, respectively; deer blood characteristics and feacal parasite egg and larval counts varied significantly between farms. Specific production outcomes chosen for risk factors analysis were; hind early conception (ie before May 1) and conception per se; hind lactational status at weaning and weaning rates (as a result of progeny losses from conception to weaning); bodyweights of weaners standardised to April 1; seasonal weaner growth rates; velvet antler grade and weight (summarised as the velvet monetary value), and the most important specific health problems, ie yersiniosis, malignant catarrhal fever and dystocia. Path analysis was the main analytical technique used in this study. Preliminary univariate analyses were carried out to identify associations between single descriptive variables and the outcome variable under investigation. Variables which showed sufficient evidence of an association in these analyses (p<0.20) were included in multivariate analyses such as multiple linear or logistic regression. A forward stepwise selection procedure was used for identification of risk factors associated with the outcome under investigation (p<0.05). Null hypothesis and final path diagrams were formulated. Final path diagrams were consistent in most respects with the current understanding of deer health an production, and allowed a more thorough investigation of causal webs within the deer production system. This study confirmed some already identified or suspected relationships, such as the positive influence of yearling hind bodyweight on conception, that of imported blood lines, wapiti blood lines and grazing swards over 5-10 cm for weaner growth, and that of stag age on velvet production. Many other relationships between risk factors and outcomes were novel, thus identifying areas for further research to identify whether they were causal. The most striking of these findings were the significant positive relationships between hind body condition scores and reproductive performance, the apparent negative effect of sunshine on hind early conception, weaner bodyweight and weaner growth, and the overall apparent beneficial effect of the presence of trees in deer paddocks on these outcomes. The other most important observation is the lack of significant risk factors predicting the probability of yearling hinds rearing a calf up to weaning, suggesting an underlying limiting physiological process for yearling hind lactation. A putative epidemiological model was proposed from risk factors analyses for management options that are likely to increase the performance. Potentially the most practical management tool defined in this study was the body condition score chart, and it is proposed that it should be used industry-wide. It is also believed a farm faecal lungworm larvae index, serum albumin, copper, phosphorus and vitamin B12 concentrations, and blood packed cell volume, haemoglobin and glutathione peroxidase activities may be useful for whole-farm investigation of health and productivity. It is proposed that final models and putative management proposals should be targeted for future research. This thesis provides the most comprehensive reference data available of farm characteristics, management practices, and health and production parameters of commercial deer farms. Data have shown a large range of productivity levels and, for the first time, potential productivity levels in a commercial farming environment have been demonstrated. These data can be used at the farm level for evaluation of performance and identification of production targets, and at the industry level for strategic planning of farm output. This study shows that there is considerable potential to improve the production efficiency and productivity on many deer farms and has identified the key factors likely to achieve that potential.