This review considers the past, present, and projected future clinical relevance of the serine protease urokinase-type plasminogen activator (uPA), and its inhibitor, plasminogen activator inhibitor-type 1 (PAI-1), in breast cancer. These factors play a key role in tumor invasion and metastasis in many cancers. In primary breast cancer, their prognostic and predictive impact has been validated at the highest level of evidence by a multicenter therapy trial (Chemo N0) and a large European Organisation for Research and Treatment Cancer-Receptor and Biomarker Group EORTC RBG pooled analysis (n = 8377). The greatest clinical use is in node-negative breast cancer, where the test can avoid over-treatment by adjuvant chemotherapy in patients with non-aggressive disease. In intermediate-risk patients as defined by the international St. Gallen consensus, it can be used to identify patients who should receive chemotherapy because their tumor is more aggressive than classical pathological factors would suggest. Gene expression signatures are already being used in clinical trials to define the population of patients with breast cancer who should receive chemotherapy. The decision for treatment ignores the highly validated information that could be provided by uPA/PAI-1. A current and future challenge is to integrate the information provided by tumor biological factors, particularly uPA/PAI-1, into refined risk assessment and decision support algorithms incorporating gene expression signatures. This article describes a paradigm ("marker fusion") for doing so and a bioinformatics approach based on this paradigm. This concept could be useful in assessing and maximizing the performance of risk assessment and the quality of therapeutic indications.