In the last few years there has been an increased interest in treatment predictive factors in breast cancer patients. The aim of the study was to analyse the role of cathepsin D and plasminogen activator inhibitor-1 (PAI-1) expression as independent prognosticators and to assess their predictive value with respect to tamoxifen treatment. This study comprises 1851 patients with primary breast cancer diagnosed during 1988-1992. Their median age was 62 years (range: 24-91). The end-point was distant disease recurrence-free interval. Adjuvant treatment with tamoxifen was given to 1136 patients (61%). The median follow-up time was 59 months (range: 39-88). Cathepsin D content was shown to be a significant independent prognosticator in multivariate Cox analysis (P=0.02). The optimal cut-off level was 10 fmol/mg protein, other cut-off levels did not improve the results. The level of cathepsin D also appeared to predict the benefit of tamoxifen among oestrogen receptor (ER)-positive patients although this result did not reach statistical significance (P=0.09). In a multivariate Cox analysis including 497 patients PAI-1 content was shown to be a significant independent prognosticator (P=0.003) but did not appear to predict the benefit of tamoxifen treatment. The optimal cut-off level appeared to be 3 ng/mg protein, which was close to the median value 2.5 ng/mg (range: 0-51). We conclude that cathepsin D is a significant independent prognosticator and may possibly also predict the benefit of tamoxifen amongst ER-positive patients. PAI-1 was also found to be a strong independent prognosticator but no treatment interaction with adjuvant tamoxifen was found.