The evaluability and reliability of proliferative activity (expressed as 3H-thymidine labeling index, 3H-TdR LI) and ploidy determinations on core needle biopsies were compared with those obtained on surgical material from the same breast cancers. The evaluability of 3H-TdR LI on core needle biopsies was markedly lower than that on surgical material (53% vs 100%), and the association between 3H-TdR LI values in the 16 cases with both evaluable determinations was poor (rs = 0.45). Conversely, determinations of ploidy on core needle biopsy and surgical material provided superimposable results, in terms of evaluability (91% vs 100%) and reliability (rs = 0.99). Further efforts are needed to improve sampling procedures for a reliable assessment of biological markers.