Many breast carcinomas are now diagnosed in needle core biopsies, after either mammographic detection or symptomatic presentation. There is dispute, however, about the range of information that should be included in the diagnostic report of these small and possibly unrepresentative samples. Is it sufficient to simply report the presence of carcinoma, in situ or invasive? Or should the histopathologist give a more detailed report including features of prognostic and predictive significance? If so, what is the evidence that the further information is, first, of clinical benefit and, second, not unreliable because of sampling variability? To address the question "What should be included in reports of needle core biopsies of breast carcinomas?" contributions were invited from authors in the USA and the UK.