Aim: Soy foods are the major source of isoflavones, which are believed to play important roles in genesis of breast cancer and its progression. We here conducted a prospective study to evaluate the association of soy isoflavone food consumption with breast cancer prognosis. Methods: A prospective study was performed from January 2004 and January 2006 in China. Trained interviewers conducted face-to-face interviews using a structured questionnaire to collect information on dietary habits and potential confounding factors. The relative risk [hazard ratio (HR)] and 95% CI were calculated from the Cox regression model for all significant predictors from cancer diagnosis to the endpoint of the study (event). Results: After a median follow up of 52.1 months (range, 9-60 months), a total of 79 breast cancer related deaths were recorded in our study, risk being inversely associated with a high intake of soy isoflavone. With an average intake of soy isoflavone above 17.3 mg/day, the mortality of breast cancer can be reduced by about 38-36%. We also found the decreased breast cancer death with high soy protein intake, with a HR (95% CI) of 0.71 (0.52-0.98). Stratified analysis with reference to the ER status, further demonstrated a better prognosis of ER positive breast cancer with a high intake of soy isoflavone (HR 0.59, 0.40-0.93). Conclusion: Our study shows the soy food intake is associated with longer survival and low recurrence among breast cancer patients. A cohort study with a larger sample size and long term follow-up is now needed.