Attending kindergarten (KG) has been shown to have persistent benefits, but the peer effects of KG remain largely unexplored. We fill this gap in the literature by using nationally representative panel data on a cohort of middle-school students (grade 7) in China. We demonstrate that when peers have had limited time to interact with one another (three to six months), there are no discernable effects of peer KG status. However, in the medium-term (14+ months), having a peer group with KG experience improves academic (math, English, and Chinese exam scores) and non-cognitive outcomes including mental health and social adjustment. These impacts are more pronounced among children from underprivileged families, and are explained by reallocation of student time and effort, a better classroom environment, improved friends' attitudes and behaviors, and pedagogically effective teacher-student interactions. The presence of these positive cross-peer spillovers indicates that the overall benefits of KG attendance are likely to be even higher than previously understood.