This presentation will report findings of a study that investigates substantive and behavioral differences between two samples in an experimental survey on consumer perceptions of foods and food labels. One sample was recruited from shopping mall visitors and the other sample from members on an established Internet consumer panel. Participants in both samples self-administered a single questionnaire on the computer. This study uses regression and other statistical approaches on paradata and survey responses to examine the influences of sample source on survey results. While there is little known literature on comparisons between these two sources of sample, existing literature on mode comparisons and on Internet surveys suggest that the two samples may behave differently and provide different survey information. Even though Internet surveys have become more widespread in recent years, mall-intercept surveys are still employed often in various research projects. This case study will contribute to the literature by increasing our knowledge about Internet surveys and about the substitutability of Internet samples for more conventional survey samples.