Non-response is a key threat to survey quality when participation is related to the topic of a survey. The higher the non-response and the more respondents differ from the non-respondents, the larger the non-response bias and the lower the validity of survey results. Refusal is usually the major cause of non-response. This can result in non-response bias when the reasons for refusal are related to the topic of the survey. Face-to-face surveys can provide some information about refusals and refusers. Based on empirical evidence from a wide range of studies this chapter shows why people cooperate and why they refuse, which survey design issues can influence cooperation and what impact interviewers can have. It also discusses the extent to which refusal conversion can help to enhance response rates and minimise bias, and how follow-up surveys or doorstep questionnaires can help to provide information about the survey questions that are central to the topic.