A common survey task is to ask respondents to recall instances that fit into a particular category. Survey questions often include cue lists—examples of items from the category being asked about. For example, in the National Longitudinal Survey of Youth (NLSY) Round 9 questionnaire, respondents are asked: Do you have any credit cards, such as Visa, American Express, or credit cards for specific stores, such as department stores or gas stations? Cue lists are thought to reduce survey measurement error by providing information for the respondent to interpret what the question is about, what types of responses should be included, and providing retrieval cues for the relevant information. However, the literature is divided on whether cues facilitate or inhibit response, whether categories (―fruit‖) or instances (―apple‖) are better, and how the effects of cues may be mediated by the question format. A web experiment was conducted in which respondents were asked questions about the convenience foods and the healthy foods they had eaten in the last seven days. The experiment varied the number of cues that respondents were presented (zero, two, or four cues) and the type of cues presented (category vs. instance). In addition, respondents were presented with one of three question formats: yes/no (Did you eat convenience foods…), frequency (How many times…), and enumeration (What convenience foods…). The findings indicate that although the effect of number of cues was slight, both type of cue and question format influenced responses. The results demonstrate how the contents of a cue list and question format can influence survey estimates and suggest implications for how survey questions should be designed.