Development of metastases in cancer patients is usually due to tumour cells seeding from the primary. These disseminated cells can be detected in blood or bone marrow with the aid of immunocytochemical methods. The clinical significance of this is, however, still controversial. With the aim of determining the prognostic value of disseminated tumour cells, a meta-analysis of 20 publications reporting results based on data collected from a total of 2494 breast cancer patients, was carried out. The outcome of the meta-analysis was that 14 of the 20 studies provided evidence suggesting positive bone marrow findings to be an unfavorable prognostic factor. Five out of twelve studies provided multivariate analytic evidence of an impact of bone marrow status on recurrence-free survival. In five out of twelve studies, a univariate analysis showed a positive bone marrow finding to correlate significantly with overall survival, while a multivariate analysis found such a correlation in only two out of six studies. Owing to a lack of standardization, the results of the individual studies are not directly comparable. Against this background it is too early to say whether the detection of disseminated tumour cells in the bone marrow can be used as a prognostic factor with an impact on therapeutic strategies.