In recent years, there has been an increasing interest in the spatial modelling of health care services: to locate services in a more equitable and efficient manner; to cope with the consequences of structural change in demography and accessibility to services; to examine the interactions between different services over space and time. In this paper, one particular spatial model used for analysing impact on patient flows, catchment populations, and other key variables resulting from changes in supply, demand, or accessibility is examined in depth. Used in several countries for planning purposes, the model was originally developed in 1979 by the Operational Research Service of the Department of Health and Social Security in the United Kingdom, for the London Health Planning Consortium. It was later refined at the International Institute for Applied Systems Analysis in Laxenburg, Austria. In the present study the robustness of prediction is tested on four versions of the basic model developed at the University of Newcastle, Australia, in 1984 and applied to data from the Sydney Metropolitan Region. Key outputs from the model are examined and the extent to which it correctly predicts changes in service levels over a three-year period in the case of five different acute clinical specialties is shown. To conclude the paper, the technical details of case-mix model are presented, which combines different acute service categories with the spatial dimension to provide a more comprehensive framework for predicting and analysing the behaviour of, and interactions between, supply and demand.