OBJECTIVES: Findings from the Women’s Health Initiative (WHI) gave rise to the concerns of risk of hormone therapy (HT). This study aimed to compare physicians’ prescribing behavior before and after the release of WHI report in Taiwan. METHODS: The data source for this study comes from the outpatient claims records of the National Health Insurance (NHI). A cohort of 57,958 women aged 40-69 who were regular estrogen users for menopausal symptom were identified during January 1, 2001 to June 30, 2001. Proportion of estrogen use, number of outpatient visits with estrogen prescription, and dosage of estrogen were compared on an annual basis, between the pre-WHI period (July 1, 2001-June 30, 2002), and the post-WHI period (Januray 1, 2003-December 31, 2003). Multiple logistic regressions and generalize estimating equation (GEE) were performed, to examine the effects of physicians’ characteristics on the changes in HT prescribing behavior. RESULTS: Probability of prescribing estrogen, number of outpatient visits with estrogen prescriptions and total dosage of estrogen prescription were significantly reduced after WHI. However, the extents to which this consulting and prescribing behavior changes varied by physicians’ characteristics: physicians of younger ages, affiliated with medical centers, and located in urban areas were less likely to prescribe estrogen, less frequently to be consulted at outpatient visits with estrogen prescription, and prescribed less dosages of estrogen, compared to their counterparts. CONCLUSIONS: The dissemination of information on drug safety seems to be varied by physicians’ characteristics. Strategies on improving the dissemination of information on drug safety,speficially targeting on less informed physician groups, were warranted to avoid the misuse of drug with safety concerns.