Many researchers have used Nationwide Food Consumption Survey (NFCS) data to estimate price and expenditure elasticities for various food groups. Results reported by these researchers have shown own-price elasticities to be negative, inelastic, and statistically significant; crossprice and expenditure elasticities have been shown to be positive and statistically significant. Further, some researchers have used NFCS data to try and estimate differential price responses of lower and higher income groups, and these authors have found no statistically significant differences in the magnitude of own-price elasticities. This study, using supermarket scanner data for several carbohydrates, finds significant differences in the price sensitivity of lower and higher income groups. These differences, however, are reflected more in purchase selections than in the magnitude of own-price elasticities. Also, regardless of income status, most expenditure and own-price elasticities are found to be elastic and statistically significant, and most cross-price elasticities are found to be positive and statistically insignificant. An important conclusion of this study is that the elasticities re-ported here provide a more accurate measure of consumer behavior than those reported in studies using NFCS data.