A model program for breast cancer screening has been initiated in Hungary. This paper evaluates the distribution of tumours on the basis of histopathological parameters in order to describe the baseline situation before the nation-wide introduction of screening. All pathology departments were asked to contribute to the study by sending in the complete anonymous reports of the last 10 breast cancer cases reported in 1997. Reports of different completeness on 419 tumours from 42 institutions (53% participation rate) were analysed. The mean age of the patients was 61.2 years (range 28-89). The mean tumour diameter was 2.64 cm (range 0.3-13). The prognostic profile (histological type, grade, lymph node involvement) of the tumours was generally poor. The information content of the reports is variable. Reports evaluable for the given factor involved prognostic information in the following proportions: tumour size: 87%; histological type: 100%; grade: 59%; skin involvement: 52%; pTNM categories; 40%; oestrogen and progesterone receptors: 37 and 29%, respectively; margins: 53%; (lymph)vascular invasion: 39%; peritumoral lymphocytic reaction: 36%; necrosis; 31%. Investigation of other prognostic markers was rare. Smaller city hospitals tended to document less prognostic factors and mostly from the less costly ones. Short and long reports showed no differences between their information content. Typewritten reports are still very common in Hungary, while computer based reporting systems requiring the filling out of prognostic markers fields are rare. The prognostic information content of these latter is generally higher. The estimated poor prognostic profile demonstrated by this report also urges the introduction of a nation-wide screening program, which should improve the observed parameters. Some degree of standardisation in pathology reporting seems also justified.