Background: It is recognized that severe HF toxicity due to TAM can compromise compliance. We previously established that HF do not correlate with endoxifen level or CYP2D6genotype, though prior reports are varied on this subject. In this pilot study, we reduced TAM dose in patients with severe HF and determined whether HF were ameliorated whilst maintaining a purported therapeutic endoxifen level of > 15nM. Methods: Eighteen patients with severe HF on 20mg TAM were enrolled. CYP2D6genotype, trough level TAM and metabolites were measured. Loprinzi HF diaries were filled out before and after DR to 10mg TAM, and hot flash scores (HFS) derived. Other data collected included demographics, smoking and alcohol history, breast cancer history, previous chemotherapies, concurrent medications, menstrual history, body mass index (BMI) and other toxicities recognized to be associated with TAM. Results: At 20mg dose of TAM, baseline endoxifen levels were 24.5, 27.9, 0-91.9 nM (median, mean, range), consistent with the high variability seen in our previous 122 patient cohort. HFS at baseline were 108, 188, 4-1482 (median, mean, range). Upon dose reduction to 10mg, endoxifen levels fell to 13.2, 18.6, 0.6-71.9 nM (difference in means p = 0.056, two-tailed T test). HFS at 10mg fell to 38, 101, 5-864 (difference in means p = 0.36, two-tailed T test). Despite this lack of statistical significance, 78% of patients reported subjective improvement of hot flashes with DR. However after DR, the proportion of patients with an endoxifen level below a suggested therapeutic target of 15nM increased from 22% to 50%. HFS did not correlate with several patient characteristics including menopausal state, duration since last menstrual period and BMI (Spearman’s rank correlation coefficient). Conclusions: In this group of women selected for having significant HF, DR of TAM from 20mg to 10mg daily resulted in halving of endoxifen levels and subjective improvement of HF, however half were below a potential therapeutic level of endoxifen at the reduced dose. Therefore routine DR to ameliorate HF toxicity may not be safe unless therapeutic drug level monitoring is performed.