Although the collection of cost and quality of life data alongside clinical studies generates detailed patient level data in a timely fashion, it also raises practical and methodological challenges. These include the fact that the settings and patients enrolled in trials may not be typical of those found in regular clinical practice, that costs and quality of life may be influenced by the trial protocol, that the clinical alternatives compared in trials may not be the most relevant for cost-effectiveness assessments, that the length of follow-up may be too short to observe changes in cost and quality of life, and that adding these data will increase the overall measurement burden in the trial. This paper discusses these challenges and the ways in which they might be overcome, focussing particularly on preference-based measures of quality of life. In particular, recommendations are given for choosing the range of quality of life instruments, sample size calculations for quality of life measurement and the measurement of quality of life in multinational studies.