Dolphins live in complex social groupings with a wide variety of social strategies. In this chapter we investigate the role that differing habitats and ecological conditions have played in the evolution of delphinid social strategies. We propose a conceptual framework for understanding natural patterns of delphinid social structure in which the spatial and temporal predictability of resources influences the ranging patterns of individuals and communities. The framework predicts that when resources are spatially and temporally predictable, dolphins should remain resident in relatively small areas. Predictable resources are often found in complex inshore environments where dolphins may hide from predators or avoid areas with high predator density. Additionally, available food resources may limit group size. Thus, we predict that there are few benefits to forming large groups and potentially many benefits to being solitary or in small groups. Males may be able to sequester solitary females, controlling mating opportunities. Observations of inshore populations of bottlenose dolphins (Tursiops sp.) and island-associated spinner dolphins (Stenella longirostris) seem to fit this pattern well, along with forest-dwelling African antelope and primates such as vervets (Cercopithicus aethiops), baboons (Papio sp.), macaques (Macaca sp.) and chimpanzees (Pan troglodytes). In contrast, the framework predicts that when resources such as food are unpredictable, individuals must range further to find the necessary resources. Forming groups may be the only strategy available to avoid predation, especially in the open ocean. Larger home ranges are likely to support a greater number of individuals; however, prey is often sparsely distributed, which may act to reduce foraging competition. Cooperative foraging and herding of prey schools may be advantageous, potentially facilitating the formation of long-term bonds. Alternately, individuals may display many short-term affiliations. These large groups make it difficult for a male or a small group of males to sequester a female, and polygynandry is the most likely mating strategy. While it is difficult to study wide-ranging delphinids to examine these predictions, this ranging and behavioural pattern has been suggested for dusky dolphins (Lagenorhynchus obscurus), coastal bottlenose dolphins (Tursiops sp.) and mixed species of dolphins in the Eastern Tropical Pacific. These patterns also resemble the ranging and social strategies of open savannah African antelopes and desert-dwelling macropods. Resource availability exists in a range of complex distributions and we predict that delphinid ranging patterns will also vary. At intermediate-ranging patterns, the framework predicts that individuals should form mid-sized groups balancing intra-group competition with predation protection. Humpback dolphins (Sousa sp.) appear to fit this pattern, with some site fidelity over relatively large ranges. They display fluid associations with other individuals. Predation pressure is not sufficiently high to cause large groups to form, and individuals probably reduce predation pressure more by hiding whenever possible. This pattern is likely to prevent the formation of long-term complex bonds. In contrast, killer whales (Orcinus orca) also display intermediate-ranging patterns, but have extremely strong social bonds within familial groups. Cooperative and altruistic behaviour in killer whales facilitate the formation of life-long bonds, similar to those observations in sperm whales (Physeter macrocephalus) and elephants (Loxodonta africana). This conceptual framework remains largely untested, and for many species it is not currently possible to describe ranging behaviours, anti-predator tactics or social behaviour in sufficient detail for appropriate examination of these ideas. Few studies on dolphins have been conducted to explicitly test this type of framework; however, existing observations of delphinid social strategies and communities are used throughout this chapter to examine this framework. Additionally, we anticipate that the present framework may provide a starting point to test hypotheses regarding the evolution of social strategies of delphinids.