The development of well tolerated, potent, specific, and nontoxic aromatase inhibitors for the treatment of postmenopausal women with estrogen-dependent breast cancer has been a major goal of recent studies. The third generation inhibitors now under investigation are nearly 10,000-fold more potent than first generation compounds. Currently available RIAs for plasma estradiol lack sufficient sensitivity to measure levels during aromatase inhibition and, thus, to assess drug potency precisely. The availability of an ultrasensitive bioassay for estradiol provided the opportunity to accurately assess the potency of a new third generation triazole aromatase inhibitor, letrozole (CGS 20267). We used this assay to measure estradiol levels in 14 women with metastatic breast cancer given letrozole at doses of 100 micrograms to 5.0 mg/day over a 12-week period. The lack of differences between doses and sampling times allowed pooling of data. Basal estradiol levels of 7.2 +/- 1.9 pmol/L (mean +/- SEM, 1.95 +/- 0.52 pg/mL) fell to 0.26 +/- 0.11 pmol/L (0.07 +/- 0.03 pg/mL) during the first 6 weeks of therapy and to 0.48 +/- 0.18 pmol/L (0.13 +/- 0.05 pg/mL) during the second 6 weeks of therapy. Although plasma estradiol levels measured by RIA were significantly correlated with levels measured by bioassay (r = 0.79; P < 0.01), the degree of suppression assessed by the bioassay (95 +/- 2% after 6 weeks) was greater than that determined by the RIA (81 +/- 4%), presumably due to improved ability to measure very low estradiol levels. We conclude that plasma estradiol is suppressed by letrozole to lower levels than previously observed, with equivalent suppression at all doses studied. A slight, although not statistically significant, rebound in estradiol levels occurs during the second 6 weeks of therapy compared to the first 6 weeks. Maximum inhibition of aromatase is achieved at letrozole doses as low as 100 micrograms.