Online shopping is common in many categories of retail goods. The recent trend towards online retailing has created an unprecedented empirical opportunity to examine consumer search behavior using click stream data. In this paper we examine consumer search intensity across a wide range of grocery products that differ in the depth of product assortment. We develop a model of attribute search in which consumers search within a chosen retailer for products that match their tastes, and that equilibrium prices reflect retailers' expectations of how intensively consumers intend to shop. The model predicts an inverse relationship between product variety and attribute search in which greater product variety reduces search intensity and leads to higher retail prices. We test these hypotheses using consumer data on online search and purchase behavior from the comScore Web Behavior Panel. Our results indicate that consumer's search less and pay higher retail prices in categories with deeper product assortments, a finding that suggests deeper product assortments can produce anti-competitive effects in retail food markets mediated through equilibrium responses in consumer search.