The incidence of breast cancer is increasing in all Western countries. Due both to a more widespread public education and to early diagnosis by mammography screening programs, the percentage of patients with node-negative breast cancer has gone up to 70%. Thus, node-negative breast cancer is a major public health problem and, consequently, clinical research in this setting is an expanding field. A recent overview analysis confirmed the results of five prospective randomized clinical trials suggesting that systemic adjuvant therapy can benefit node-negative breast cancer patients. Because of the heterogeneity of node-negative breast cancer, it is reasonable to attempt to avoid excessive treatment morbidity and costs by using selective prognostic markers to identify patients at high risk for disease recurrence who are eligible for postsurgical systemic adjuvant therapy. It is also desirable to use predictive markers in selecting the therapy to which each patient is more likely to respond. The need for additional prognostic and predictive factors has led to identification of a plethora of potentially useful markers. As a result, the selection of patients at different risks of developing node-negative breast cancer and the choice for appropriate therapy remain difficult and confusing for the clinician. Moreover, the majority of studies have examined new markers individually rather than by multivariate analysis and retrospectively rather than prospectively. Thus, there are also important methodologic biases in such studies. This analysis consists of (a) defining the clinical "problem," (b) defining the terms of prognostic and predictive factors, (c) suggesting more appropriate laboratory and clinical approaches to properly evaluate a new indicator, (d) identifying the subsets of patients in whom the use of new prognosticators is warranted and of particular importance, and (e) providing some direction for future research on this topic. Our ultimate goals are to facilitate the understanding of node-negative breast cancer prognostic markers among clinicians, to help them select the most appropriate indicator for specific situations, and to recommend methodology for future research.