The addition of lapatinib (Tykerb/Tyverb) to capecitabine (Xeloda) delays disease progression more effectively than capecitabine monotherapy in women with previously treated HER2+ metastatic breast cancer (MBC). The quality-adjusted time without symptoms of disease or toxicity of treatment (Q-TWiST) method was used to compare treatments. The area under survival curves was partitioned into health states: toxicity (TOX), time without symptoms of disease progression or toxicity (TWiST), and relapse period until death or end of follow-up (REL). Average times spent in each state, weighted by utility, were derived and comparisons of Q-TWiST between groups performed with varying combinations of the utility weights. Utility weights of 0.5 for both TOX and REL, that is, counting 2 days of TOX or REL as 1 day of TWiST, resulted in a 7-week difference in quality-adjusted survival favouring combination therapy (P=0.0013). The Q-TWiST difference is clinically meaningful and was statistically significant across an entire matrix of possible utility weights. Results were robust in sensitivity analyses. An analysis with utilities based on EQ-5D scores was consistent with the above findings. Combination therapy of lapatinib with capecitabine resulted in greater quality-adjusted survival than capecitabine monotherapy in trastuzumab-refractory MBC patients.