Open-ended questions can provide valuable information to help researchers understand a respondent’s thinking but using such questions has proven problematic in mail surveys. Experimental data are used to assess the impact of answer space size on the length and content of responses to two open-ended questions that are part of Florida Cooperative Extension’s customer satisfaction survey. Based on nearly 2,000 responses collected from 2003 to 2006, the size of the answer space showed no affect on the propensity to answer, but did impact the length of the response with larger spaces eliciting more words and sentences. Results of the content analysis will also be presented. These data provide a context for researchers to make decisions about the use of open-ended questions in mail surveys and the potential impact on the information collected.