A population-based study on the therapeutic effects of a diabetes teaching programme (DTP) based on problem oriented participatory education (POPE)--a method based on learner activity in group meetings--was undertaken at the Primary Health Care Centre, Kisa, Sweden, in collaboration with educationalists. A control group was given conventional classroom teaching. To be included a patient had to be aged 55-73 years, to live in his own home, and to have non-insulin-treated type II diabetes mellitus. The therapeutic effects of the DTP were studied before, during, and after POPE with regard to three factors, diabetes related knowledge, behaviour assessed by dietary and exercise habits, and the quality of the anti-diabetic therapy as assessed by metabolic profile including Hb-A1. Significant improvement in knowledge and transient improvement in Hb-A1 concentration were recorded among patients taking part in a DTP adjusted to their individual problems and needs. When improvement in metabolic control does not last, group meetings should be continued for more than the three months used in the present study. We believe that such improvement is intimately bound up with the psycho-social process that is involved in the group meetings and that helps the patient to cope with the disease in particular and life in general.