Commodity prices have been rising at unprecedented rates over the last two years. The primary objective of this paper is to assess if and how firms pass through upstream cost increases to final good prices. First, we investigate what happens to the shelf prices (the regular prices) of goods that contain significant amounts of a commodity whose price has changed. The objective is to document patterns of price rigidity depending on the share of the commodity in the final good that is sold to consumers. For example, given an abnormal commodity price change in wheat, what happens to the shelf regular price of bread, wheat cereals, and other goods that contain wheat? Commodity pass-through patterns for ready to eat cereal (smallest share of commodity in final product) and fresh chicken (largest share of commodity in final good) are investigated. Second, we also assess what happens to the net prices consumers pay (that is the regular price net of discounts offered). One possible way to pass through a cost increase is to reduce the frequency of promotional discounts, or offer smaller discounts to consumers. Upstream commodity input prices used in our investigation are wheat and corn futures prices, to account for upstream inputs, and flour and chicken feed producer price sub indices for downstream cost shocks. We combine several datasets for this empirical analysis: commodity prices, commodity price indices, and scanner data on prices for a variety of goods, over a four year time period and across several stores in California, belonging to a large retail chain. We construct quantity weighted price indices within two product categories sold in the supermarket, where prices are weighted by pre-determined quantity weights to obtain shelf price indices and net price indices. For each of the commodities, regressions will be run using store-level product (UPC) weekly data. The reduced form regressions consist of projecting the shelf price index, as well as the net price index, on commodity prices, other explanatory variables and on region and time dummies. The point estimates measure the effect of residual changes in commodity prices, net of seasonal and regional effects, on the prices consumers face when making purchase decisions. We also construct a variable that measures the frequency of price discounts and relate that variable to the same explanatory variables. We estimate pass through behavior using the above three different measures of retail price activity controlling for cost proxies, store-level fixed-effects and regional time trends using panel data estimation techniques. Results suggest that an important part of retail price variation comes from promotional activities, and the usual shelf price index would underestimate the true pass through coefficient. To deal with omitted variables and price stickiness we included a lagged dependent variable, using the Arellano-Bond dynamic panel estimator. For Chicken the results show that using standard information on regular shelf price leads to an underestimation of the true pass-through coefficient. For Cereal, using standard shelf prices leads to an overestimation of the pass-through coefficient reflecting the importance of storability faced by consumers and retailers, and industry characteristics in the sale dynamics. Not only do our cost pass-through estimates account for sales we also provide dynamic multipliers for grain commodity price increases to supermarket shelf prices. The estimated dynamic elasticities are not as small as one might expect from a naive model. The elasticity of cereal price with respect to flour is over 1 and the elasticity of chicken price with respect to chicken feed was 30 percent. These estimates would imply a very large price increase in cereal and chicken over the last several years. There are fewer sales when commodity prices go up. From this we would conclude that net prices should be used for pass-through analysis.