BACKGROUND: Effective management of breast cancer is dependent on adequate pathological reporting of the surgical specimen. OBJECTIVE: To describe the frequency with which histopathological features of known prognostic importance are routinely recorded. STUDY POPULATION: 885 cases of invasive breast cancer diagnosed in NHS laboratories in Lancashire and Greater Manchester. METHODS: Pathology reports were reviewed for details for tumour histological type, size, and grade, the presence or absence of tumour in blood or lymphatic vascular channels, and a comment on the proximity of tumour to the lines of surgical excision. Laboratories were categorised according to their throughput of cases of breast cancer, involvement in the breast screening programme, and whether they were attached to a teaching hospital. RESULTS: Histological type, tumour size, presence or absence of tumour in vascular channels, and adequacy of excision were recorded for 843 (95%), 803 (91%), 436 (49%), and 761 (86%) cases, respectively. Non-screening and low throughput laboratories were significantly less likely to record certain histopathological features. No significant differences were observed between teaching and non-teaching hospitals. CONCLUSIONS: The substantial interlaboratory variation in the histopathological reporting of breast cancers can, in part, be related to throughput of cases and involvement in the breast screening programme.