Urokinase-type plasminogen activator (uPA) may be responsible for the invasive and metastasizing capacity of tumor cells. Evidence has been presented that primary breast cancer patients with tumors containing high levels of uPA experience a worse prognosis. In the present study we have assessed uPA status in routinely prepared cytosols of 671 primary human breast tumors and have evaluated its association with disease-free and overall survival. Isotonic regression analysis with length of disease-free survival as an end point revealed 1.15 ng/mg protein as the best cutoff point to discriminate between uPA positive (32% of the tumors) and uPA negative. In both Cox univariate and multivariate regression analysis (including also patient's age, menopausal status, lymph node status, and the number of positive lymph nodes, tumor size, and estrogen and progesterone receptor status), uPA positivity was significantly associated with increased rates of relapse and death. Corrected for all relevant factors in multivariate analyses for subgroups of patients, uPA positivity was significantly associated with an increased relapse rate in the subgroups of node-negative (P = 0.002; relative failure rate, 2.33), node-positive (P < 0.0001; relative failure rate, 1.95), postmenopausal (P < 0.0001; relative failure rate, 2.59), and steroid receptor-positive patients (P < 0.0001, relative failure rate, 2.76). We conclude that uPA positivity of human primary breast tumors is an important independent variable for the identification of patients at high risk for recurrence, also in clinically important subgroups of patients.