AIMS The purpose of the study is to evaluate histological measurement methods for quantitative assessment of the degree of tubular differentiation in breast cancer. METHODS AND RESULTS We evaluated tubular differentiation in 20 cases of invasive breast cancer by four different assessment methods. Method 1 was the traditional subjective evaluation of the amount of malignant tubules in each sample. Method 2 evaluated the fraction of fields presenting tubular differentiation by registering the presence or absence of neoplastic tubular structures in each microscopic field. In method 3 the area fraction of malignant epithelial cells presenting tubular differentiation was assessed field-by-field and expressed as an average of the whole tumour area. Method 4 applied point counting for evaluating the fraction of malignant epithelial cells in tubular structures. By correlation and reproducibility analyses, method 1 was inferior to the other methods. Method 4 was accurate but too laborious and time-consuming for clinical use. Methods 2 and 3 were both efficient and reproducible and could be used interchangeably. With the time and effort used in the measurements taken into consideration method 2 was best applicable to clinical practice. CONCLUSION Accurate evaluation of tubular differentiation in breast cancer is possible by defining the presence or absence of tubular differentiation in microscopic fields of a histological section. Assessment of the fraction of fields with tubular differentiation (FTD) is simple, unambiguous, objective and fast--even a large sample can be screened in less than 10 min. In our results, FTD has clear advantages over subjective or point counting-based evaluation methods of tubular differentiation.