A good questionnaire design for a clinical trial will minimise bias and maximise precision in the estimates of treatment effect within budget. Attempts to collect more data than will be analysed may risk reducing recruitment (reducing power) and increasing losses to follow-up (possibly introducing bias). The mode of administration can also impact on the cost, quality and completeness of data collected. There is good evidence for design features that improve data completeness but further research is required to evaluate strategies in clinical trials. Theory-based guidelines for style, appearance, and layout of self-administered questionnaires have been proposed but require evaluation.