Most of us are busy in our work – patients to see, reports to write, meetings to attend, students to teach etc. Then another questionnaire crosses our desk, sometimes prettied up, e.g. with a tea-bag stapled to it so you can complete the survey ‘over a cuppa’ or a ‘smiling face’ sticker attached or the offer of money on the questionnaire's return. The second last thing you feel like doing is completing the wretched thing, whether over a cup of tea or something stronger. The very last thing you feel inclined to do is conducting a survey yourself. Nevertheless, surveys can provide valuable information for consumers, clinicians, administrators and policy-makers. Surveys are often more popular to carry out than other studies and, for would-be researchers, can be a gentle introduction to research. Nevertheless, conducting a survey does not amount to thinking up a list of questions on a rainy day and inflicting them on your colleagues; skills and etiquette are required. As with all research, there are problems and pitfalls. Here is a step-by-step guide, based on the authors' experience [e.g. 1–6] and reading [e.g. 7–12], which should make your survey easier to conduct, and more palatable for the would-be respondent.