Proliferating cell nuclear antigen, PCNA (PC10), immunolabelling was determined in 175 women with breast carcinomas and related to other established prognostic factors: flow-cytometric data, volume-corrected mitotic index, sex steroid receptor content and clinical outcome during the mean follow-up of 9 years. The maximum fraction of PCNA-positive nuclei (PCNAmax), the average fraction of positive nuclei (PCNAtot) and the number of intensely stained nuclei per microscope field (PCNAcount) were significantly related to histological grade (P < 0.001), DNA ploidy ((P < 0.001), S-phase fraction (P < 0.001), mitotic index (P < 0.001) and sex steroid receptor content (P = 0.002). PCNAmax (P = 0.015) predicted survival in univariate analysis; PCNAtot (P = 0.025), PCNAmax (P = 0.007) and PCNAcount (P = 0.019) predicted the recurrence-free survival. In axillary-lymph-node-negative tumours, PCNAtot (P = 0.092), PCNAmax (P = 0.036) and PCNAcount (P = 0.006) predicted survival and recurrence-free survival (P = 0.011), (P = 0.012) and (P = 0.006) respectively. In multivariate analysis including clinical, histological, flow-cytometric and biochemical variables, PCNAtot (P = 0.004) predicted the recurrence-free survival independently. In axillary-lymph-node-negative breast cancers, PCNAtot predicted accurately the patient survival (P = 0.002) and the recurrence-free survival (P = 0.002). The results indicate that PCNA immunolabelling has independent prognostic value particularly in local breast cancer.