This paper presents a summary of the evidence review group (ERG) report into the evidence for the clinical effectiveness and cost-effectiveness of gemcitabine with paclitaxel for the first-line treatment of metastatic breast cancer (MBC) in patients who have already received chemotherapy treatment with an anthracycline, compared with current standard of care, based upon the manufacturer's submission to the National Institute for Health and Clinical Excellence (NICE) as part of the single technology appraisal (STA) process. The clinical evidence for gemcitabine as a treatment for MBC comes from the unpublished JHQG trial (some data commercial-in-confidence): overall survival was 3 months longer for the gemcitabine/paclitaxel arm (18.5 months) than for the paclitaxel arm (15.8 months) (p = 0.0489); gemcitabine/paclitaxel also improved tumour response and time to documented progression of disease compared with paclitaxel monotherapy, but haematological serious adverse events were more common. In the absence of any formal methods of indirect comparison there is insufficient robust evidence to compare the relative effectiveness of gemcitabine/paclitaxel with docetaxel monotherapy or docetaxel/capecitabine combination therapy. The manufacturers used a Markov state transition model to estimate the effect of treatment with five different chemotherapy regimes, adopting a 3-year time horizon with docetaxel monotherapy as the comparator. Health state utilities for different stages of disease progression and for patients experiencing treatment-related toxicity are used to derive quality-adjusted life expectancy with each treatment. The base-case cost-effectiveness estimate for gemcitabine/paclitaxel versus docetaxel is 17,168 pounds per quality-adjusted life-year (QALY). When longer survival with docetaxel is assumed in a sensitivity analysis, the incremental cost-effectiveness ratio (ICER) is 30,000 pounds per QALY. Probabilistic sensitivity analysis estimates a 70% probability of gemcitabine/paclitaxel being cost-effective relative to docetaxel at a willingness-to-pay threshold of 35,000 pounds. There is considerable uncertainty over the results because of the lack of formal quality assessment or assessment of the comparability of the 15 trials included in the input data, and the questionable validity of the indirect comparison method adopted. An illustrative analysis using a different method for indirect comparison carried out by the ERG produces an ICER of 45,811 pounds per QALY for gemcitabine/paclitaxel versus docetaxel. The guidance issued by NICE in November 2006 as a result of the STA states that gemcitabine in combination with paclitaxel, within its licensed indication, is recommended as an option for the treatment of MBC only when docetaxel monotherapy or docetaxel plus capecitabine is also considered appropriate.