In many countries manpower problems in the field of health care are regular items on the agenda of policy makers. To avoid mismatches between demand of care and supply of care on national and regional levels, manpower planning models and methods are used to determine adequate numbers of medical specialists to fulfil the future demand of care. A key question is which factors should be included in long term (25 years) forecasting and how these factors should be included in a planning model and method. In this paper we evaluate the model and the method for medical manpower planning that has been used since 2000 in the Netherlands. To improve accuracy and relevance of the model, we conclude that system dynamics modelling should be used and that strict distinction should be made between factors influencing the demand for care and factors influencing the supply of care.