How do migrant social networks matter for performance in the job? We examine this by constructing a nationality-based network of foreign newcomers when they first begin to play in the PGA TOUR and examine the impact of this initial social network on newcomers’ probability of surviving (i.e., keeping their license) at the end of their inaugural PGA TOUR season. We find that the migrant social network matters among the non-elite group of players in the second tier of the PGA TOUR, but not among the elite group of players in the first tier of the PGA TOUR. For the second-tier tour players, we find that density of connections within a nationality cluster has a sizable positive effect on newcomers’ probability of surviving, but no evidence that the centrality of a nationality cluster in the overall PGA TOUR network has an impact on survival.