The scale of human cooperation and conflict is outstanding and evolutionarily challenging to explain. Cooperative and hostile behaviours have deep evolutionary roots and adaptive functions. However, theoretical models differ in how they explain these functions. Thus, my thesis aims to empirically test functional predictions about human cooperation and conflict. These experiments use a behavioural ecological framework, and pay also attention to the effects of social and developmental environments. In the first two chapters, I studied how group composition affects cooperation and individual success, and how people react to information of each other’s cooperative behaviour. I found that cooperative contributions increased with greater group heterogeneity for those with high baseline cooperativeness, and decreased for those with lower baseline cooperativeness. However, people were insensitive to pre-information of group composition, even though group composition was essential for cooperation to be successful. In Chapter III, I compared evolutionary theories of human intergroup conflict by empirically testing whether intergroup aggression is motivated by public goods as predicted by models of group selection (i.e. Parochial Altruism models) or by private goods as predicted by models of individual selection (i.e. Male Warrior and Chimpanzee models). My results challenge group selection models by implying that private goods motivate intergroup hostilities even when they are socially inefficient. In chapter IV, I propose and test a hypothesis that harsh parental treatment intensifies group-beneficial prosocial and bellicose norms and thus contributes to group success in intergroup conflicts. I found, in accordance with my hypothesis, that harsh parental treatment increased both these traits in males and harsh parenting may thus bear cultural evolutionary consequences. Together, my results contribute further knowledge to our evolutionary understanding of the dynamics, backgrounds and patterns of human cooperation and conflict.