It is claimed that high returns can be achieved from hunting and ecotourism operations. As a result wildlife production is a rapidly growing form of land-use in South Africa. Lately, rural African communities have approached regional conservation agencies for aid to establish small game reserves so that they too may benefit from wildlife production. However wildlife operations have high input costs relative to domestic stock operations and no attempt has been made to determine the effect of property size on the costs and revenue generated by wildlife. It is thus necessary to conduct a Cost-Benefits Analysis to ascertain this effect by determining the opportunity cost incurred by choosing wildlife over other land-uses suitable in semi-arid savannas, namely communal subsistence production and commercial beef production. This project attempts to quantify the revenue generated, and the variable costs and fixed costs incurred by wildlife production, subsistence production and commercial beef production in order to observe their behaviour against property size and by this means to establish the size ranges for which each of the three land-uses is most appropriate. Mathematical modelling is used to define each of the three land-uses and how their revenue and cost curves interact with property size. The resultant profit curves are able to assess only the financial benefits from each of the land-uses to the local community. An assessment of the full economic benefits to the local and broader community would require different criteria and apportionment of costs and revenue. The effect of property size on fixed costs is the single most important factor which distinguishes the behaviour of the profit curves of the three land-use options: subsistence production has negligible fixed cost input and so is able to achieve greater profitability than either beef or wildlife at small property sizes. Beef has high input costs per hectare at small land sizes which diminish with each unit of additional land. Wildlife operations also have high input costs at small land-sizes which decrease per hectare with additional land added. However due to the service industry nature of wild life operations, fixed costs increase per hectare after some point (in this case it is assumed to be 2000 ha). This is because the attractiveness of game reserves to tourists increases with size due to the inclusion of "many" species of game, which in turn increases the number of people entering the park per hectare and as such the fixed cost input required to accommodate those extra people. The specific results derived from the model indicate that the profit curve of wildlife rises far more steeply than those of either subsistence production or commercial beef production. However, due to the effect of input costs, both commercial beef and subsistence production are more profitable at land sizes of less than 3000 ha. This indicates that investing large sums of money into small game reserves of less than 3000 ha may not be justified on the basis of profits alone. / Thesis (M.Sc.)-University of Natal, Pietermaritzburg, 1998.