Determinants of periodontal treatment costs are explored by constructing and estimating a recursive multi-equation model. The costs are based on treatment needs defined by the conventional concept of periodontal disease. Special attention is paid to the effect of cigarette smoking on the costs. The model is based on the health production theory and the findings of earlier studies of factors affecting periodontal disease. The model was estimated by Lisrel by using the ML method. The data were based on a clinical examination and interview of a representative sample ( n = 4771) of the dentulous Finnish population aged 30 years and over (The Mini-Finland Oral Health Survey). Most of the predictions of the model were confirmed by the empirical testing. In terms of direct effects, there seems to be four significant explanatory factors to the total costs: the number of remaining natural teeth, age, caries status and smoking in that order. The elasticity of costs with respect to the number of cigarettes per day was 0.072 for men and 0.021 for women. Moreover, the model revealed other interesting connections between oral health variables, oral hygiene and health care habits and the exogenous variables of age, education and income. Implications of new concepts of periodontal disease on results are also discussed.