With the rapid development of e-commerce, online shopping behavior gets more and more attention. Consumers can justly stay at home to enjoy shopping online. Shopping online not only saves a lot of time and energy costs, but also improves the possibility of consumers finding products to meet their needs. Sometimes consumers get benefit from plentiful options, sometimes they spend a lot of time and effort but cannot find the satisfaction to meet their needs. Results of previous studies are inconsistent. A lot of studies shown that the more the better, however, some researchers found “choice overload effect”. Choice overload effect states that when the individual is faced with too many options, compared to a smaller set of options will lead to more negative results. For example, the lower degree of satisfaction, delaying selection, regret improved, or the decline in self-confidence, buying behavior reduction. A large number of studies in the past were discussed, but the majority were on traditional shopping. The empirical research about assortment size and choice overload effect of online shopping is few. The present study aims to explore the effect of assortment size, target certainty and product type on consumer decision-making attitude and behavior for online shopping. Three pretests and a main laboratory experiment were conducted. First, we ran pretest 1 and 2 to select the specific product materials to be used in the main experiment, which would be used to indicate main attributes for experience product and search product (beef jerky as experience product and USB flash disk as search product). Second, we ran pretest 3 to make sure the option number for big and small selection set (15 options as small selection set and 60 options as large one). 125 students (41 for pretest 1, 54 for pretest 2 and 30 for pretest 3) participated in the pretests. It used a 2(large, small) ×2(certainty, uncertainty) ×2(experience product, search product) between-subject design in our main experiment. The participants’ target certainty was manipulated by whether the participants needed to describe what they want. A total of 160 students (81 males and 79 females, mean age=21.89, SD=3.70) were randomly assigned to eight groups. They were first asked to read the instruction and imagine that they were in online shopping. Then they were manipulated of target certainty and scanned the webpages we designed for each group. Finally, they completed a series of questions measuring their satisfaction, confidence and choice deferral. The results showed the attitude-behavior separation, choice from large assortments were more likely to lead to higher satisfaction (M=5.16 vs. 4.55), higher confidence (M=5.11 vs. 4.73) but more choice deferral (78.8% vs. 58.8%). Target certainty moderated the effect of assortment size on choice deferral, target uncertainty individuals showed more choice deferral in the large assortments than in the small assortments (82.5% vs. 50.0%). While target certainty individuals had similar choice deferral between different assortment size. Product type didn’t moderate the effect of assortment size on choice deferral. Comparing with prior studies on assortment size, our results suggested that: in the online shopping environment, in terms of the attitude level, consumers still preferred large selection set, and the select-satisfaction and the self-confidence were higher than in the small selection set; however, in terms of the behavior level, in big choice set there were more negative buying behavior and consumers preferred to delay purchase. This study explored the choice overload effect and attitude-behavior separation on e-commerce, which had implications for webpage layout.