My PhD project intends to explore the importance of play in child development. To examine the longitudinal influence of play on children's learning and mental health development, I conduct three studies by using the Longitudinal Study of Australian Children population cohort and the National Education Panel Study (in Germany). These three studies all intend to understand whether children’s play experiences and skills at around age 2/3 would affect their language, mathematical and mental health outcomes at around age 6/7. My study one explores the relation between symbolic play skills and structural language development among autistic and non-autistic. Through propensity score matching, I matched 92 autistic children and their 92 non-autistic peers in the Longitudinal Study of Australian Children population cohort. These two groups of children had exact same sociodemographic backgrounds and development levels. Through correlational and hierarchical regression analyses, I discovered that autistic children’s symbolic play skills at age 3 significantly contributed towards their semantics, syntax, and coherence skills at age 7. My study two explores the role of playful home mathematics environment activities on children’s mathematical development. I used structural equation modelling and mediational analyses with bootstrapping on 1,184 children in the National Education Panel Study. My study two shows that a mathematically and linguistically stimulating, attentive and responsive home learning environment at age 2, alongside with more frequent formal and informal home mathematics at age 5 were essential for children’s mathematical development between age 4 and 6. My study three explores the role of peer play abilities in children’s mental health development. I conducted analyses on 1,676 children from the general population in the Longitudinal Study of Australian Children population cohort, as well as on 243 children from the high-reactivity group and 270 children from low-persistence group, who tended to be at higher risks for mental health problems. Through structural equation modelling, I demonstrated that better peer play abilities at age 3 predicted lower risks of internalising and externalising problems at age 7 for the general population. Better peer play abilities at age 3 also predicted lower risks of hyperactivity problems for the high-reactivity group, whereas better peer play abilities at age 3 predicted lower risks of hyperactivity, emotional and peer problems for the low-persistence group. These studies all together suggest that play in early years could generate positive growth in children’s language, mathematical and mental health development, for both neurotypical and neurodivergent children. Educational and practical implications are provided, alongside with suggested future directions.