My dissertation focuses on the intersection between immigration and social networks. The three essays of the dissertation study network formation, network characteristics, and network effects, respectively. The first essay investigates how immigrants construct carpooling networks in order to deal with language problems when commuting. I focus on the role of language proficiency, and find that immigrants with lower levels of English skills are more likely to commute to work by carpooling. Similarly, the number of co-riders is negatively associated with English proficiency. In other words, immigrants create need-based carpooling networks in order to tackle potential language problems. The second essay studies how social networks can be defined based on a typical acculturational behavior, namely, English-name usage. Exploiting a natural linguistic experiment, I find that Chinese students with English-name usage have more close friends who are also English-name users. This implies that homophily could occur among friendships within the same ethnic group in the context of immigration. The third essay examines social network effects among highly professional migrants: I focus on French football players in England and study whether ethnic networks affect yearly migration outcomes. I find that a player exposed to a larger French network is more likely to stay in England, although not necessarily the same team. However, the network effects are highly heterogeneous, and ethnic networks do not always benefit those who need support most, such as veteran players or players with relatively low levels of outputs.