Several studies have shown that hormonal, metabolic and inflammatory mechanisms may affect breast cancer progression. We tested the prognostic value of metabolic syndrome in 110 postmenopausal breast cancer patients, who participated in a 1-year dietary intervention study. The risk of adverse events after 5.5 years of follow-up was examined by Cox' proportional hazard modelling, adjusting for hormone receptor status, stage at diagnosis and serum testosterone level, which were shown to significantly affect prognosis. The adjusted hazard ratio of recurrence for the presence of metabolic syndrome at baseline was 3.0 (95% CI 1.2-7.1). Combining metabolic syndrome and serum testosterone, the adjusted hazard ratio of recurrence among women with metabolic syndrome and testosterone levels higher than 0.40 ng/ml (median value) was 6.7 (95% CI 2.3-19.8) compared with that among women without metabolic syndrome and testosterone levels < or =0.40 ng/ml. The results suggest that metabolic syndrome may be an important prognostic factor for breast cancer.