Immunohistochemical (IH) assessment of nuclear estrogen receptor has been considered an alternative method to conventional biochemical assay. The present work intends to compare specificity and sensitivity of IH and biochemical technique to assess nuclear estrogen receptor in formalin-fixed and paraffin-embedded mammary carcinoma samples. IH positive reaction was defined as 14% or more nuclear staining in 100 cells counted under high magnification (400x). Biochemical assay was considered positive over 10 fmol/mg of protein. 66 cases were collected with a mean age of 55.6 years and a mean tumor size of 25.2 mm. Histologically, 62 cases were ductal carcinomas, 2 lobular carcinomas, and 2 medullary carcinomas. Biochemical assay for estrogen receptor was positive in 35 cases (63%) and IH in 40 cases (71%). The present results show that IH assessment of estrogen receptor is highly specific and sensitive. Estrogen receptor present in non-tumor cells and blood vessels walls may disclose false positive biochemical results and false negative result if the tumor mass is small or there are isolated tumor cells. IH assessment of estrogen receptor can be performed in small samples, including in situ lesions. The method of fast, reliable, and of lower cost, IH may be considered the method of choice in cases with insufficient sample for biochemical assay and/or in tumors containing scant cells.