Most of the research concerned with the economics of health systems has focussed on allocative efficiency. Specifically, much effort has been devoted to the development and application of techniques of economic evaluation. The consideration of technical efficiency has figured less prominently in the search for 'solutions' to the problems of health systems. Those working on the economic evaluation of health care interventions have adopted the assumption that interventions are being, or will be, produced in a technically efficient manner. The aim of this thesis is to challenge this assumption and illustrate the potential implications of assuming technical efficiency when allocating scarce resources. Two case studies from Bangladesh are presented: vaccination services in Dhaka City and primary health care in rural Bangladesh. The specific objectives of this thesis are to: estimate the cost of these services using standard costing methods; and analyse the same data sets using parametric (stochastic frontier analysis) and non-parametric (data envelopment analysis) techniques in order to identify whether, and to what degree, the services were being delivered efficiently. Applying efficiency measurement techniques illustrated that standard costing methods disguise a high degree of inefficiency. By investigating production practices, costs related to inefficiencies can be identified and addressed. The thesis illustrates that if something is deemed worth doing then it should be carried out in a way which ensures the optimum use of scarce resources. An exclusive focus on switching resources from less cost-effective to more cost-effective activities will not realise the full benefits in terms of improved allocative efficiency if providers on the ground are not producing services at lowest cost. Recommendations are made for policy-makers on how technical efficiency can be improved. Recommendations for future research are also made.