The booming of E-commerce trading sites, especially after COVID-19, has aggressively put pressure on the competition in this industry. Hence, building young customers’ loyalty plays a vital role in doing business. One of the best solutions to this problem is to boost customer satisfaction. As customers are satisfied with the products or services provided, they tend to purchase more frequently and eventually become more loyal. Regarding location, HCM City is currently the most potential market for E-commerce development. Therefore, the aim of this paper is to weigh the relative impacts of customer satisfaction factors on the youth’s loyalty (18-25 years old) to E-commerce trading sites in HCM City, address the existing issues, and propose the corresponding solutions. The study used Multiple Linear Regression Analysis with SPSS 20.0 to build, examine, and test the research model through 215 samples indicates that there are five factors, which affect the youth’s loyalty. They are the satisfaction of website design, the satisfaction of the product portfolio, the satisfaction of privacy & security, the satisfaction of delivery, and the satisfaction of the post-sales service. The results reveal that the satisfaction of privacy & security is the most influential factor in the youth’s loyalty, followed by the satisfaction of the post-sales service and the satisfaction of website design. In addition, there are no differences in young customers’ loyalty regardless of genders, occupations, frequencies of visiting E-commerce sites, monthly income levels, and average spending per order. Accordingly, these findings can be good guidance for E-commerce trading sites to formulate suitable development strategies to build the loyalty of customers at the early stage (18-25 years old) and effectively compete in the market. On the other hand, the paper has certain limitations, namely the short duration of collecting primary data (one month) and thus the relatively small sample size compared to the whole population of the youth. Moreover, as the method of data collection is survey-based, the study must be remade for different regions and nations. Also, this paper only covers E-commerce trading sites within the retailing industry. Hence, future research can greatly improve the model by extending the time period of data collection and covering more industries. Besides, a more modern approach such as SEM is a decent option for measurement scales evaluation and model testing, since Multiple Linear Regression Analysis requires strict assumptions and may have a great likelihood of measurement errors.