OBJECTIVES: To evaluate cost effectiveness (CE) of Paclitaxel compared to Docetaxel for Anthracycline pretreated metastatic breast cancer (MBC) patients from a societal perspective. METHODS: A two year decision model was developed with parameter inputs from published literature and the first phase III randomized clinical trial which compared the taxanes for treatment of MBC. Direct cost (in 2005 $) included in the model were drug and premedication cost, administration and personnel cost, cost of hospitalization and adverse effects, laboratory cost, home health aide cost along with follow up and terminal care cost. Indirect cost consisted of informal caregiver time and patient time cost. Effectiveness was measured in terms of quality adjusted life years (QALYs), which was based on utilities elicited from US oncology nurses and life expectancy calculated using declining exponential approximation of life expectancy (DEALE) method. All costs and QALYs were discounted at 3% and an incremental cost effectiveness ratio (ICER) was calculated. Uncertainty of point estimates was analyzed in Univariate sensitivity analysis. Threshold sensitivity analysis was also conducted to evaluate the values where the CE ratio changed. RESULTS: Paclitaxel was more cost effective for treatment of MBC, amongst the two taxanes. Docetaxel drug cost and adversity profile made the ICER $145,837/QALY gained. The model was robust to reasonable changes to the parameter estimates. Response to treatment was one of the key parameters which affected the CE ratio. Threshold analysis suggested, either Docetaxel response rate should increase to 42% or Paclitaxel response rate decrease by a third or the price of Docetaxel decrease by one sixth, to bring the ICER close to the $100,000/QALY. CONCLUSIONS: Docetaxel may be a better choice from clinical stand point for treatment of MBC patients, but its economic justification is questionable in a cost conscious society with limited resources.