Data from 2933 consecutive cases of primary breast carcinoma, observed in our Institute from 1984 to 1994, having documented estrogen (ER) and progesterone (PR) receptor levels, were obtained from the Institute's Hospital Tumor Registry and analysed after being categorised as follows: age, less than or equal to 60 vs. >60; menopausal status, pre-menopausal vs. post-menopausal; histology, ductal vs. lobular vs. others; tumor size, T-1 vs. T-2, T-3, T-4; nodal status, N-0, vs. N+; histologic grade, 1-2 vs. 3; focality, unifocal vs. multifocal; ER status, <10 fmol/mg protein vs. greater than or equal to 15. At multivariate analysis, using a logistic model including age, histology, tumor size, nodal status, histologic grade, uni-multifocality and PGF/ER status, significant associations were, for ER status: PGR status (OR = 34.01, 95% CI:20.08-57.80), histology (OR = 3.24, 95% CI:1.85-5.67), histologic grade (OR = 2.18, 95% CI:1.38-3.42), menopausal status (OR = 2.17, 95% CI:1.26-3.74), age (OR = 34.01, 95% CI:20.08-57.80), menopausal status (OR = 5.27, 95% CI:1.43-3.33), age (OR = 1.71, 95% CI:1.13-2.59). The finding that estrogen receptor positivity was more prevalent among tumors with lobular histology seems to suggest the possibility of fundamental differences in tumor biology ductal and lobular cancers.