It is believed that the results of the marshmallow test predict important outcomes throughout the life span. However, this is not entirely certain. Individuals vary in their ability to regulate behavioural, emotional and attentional impulses to achieve long-term goals. These individual differences show stability across the lifespan. This means that the marshmallow test may be predictive of later outcomes because the self-control remains stable throughout life. The marshmallow test may also be predictive because it captures social and situational factors. People are less willing to delay gratification with untrustworthy individuals or when delay is not normative or rational. A child’s ability to wait during the marshmallow test may thus be less important than the social and environmental circumstances influencing their willingness to wait.Pre-schoolers who delayed gratification had significantly better academic achievement (1), fewer problem behaviours (2) and better social skills (3) in adolescence than those who did not delay gratification. Models that adjusted for pre-school levels of the adolescent outcome constructs continued to predict significantly less problem behaviour in adolescents. The relationships between delay of gratification and academic achievement and social skills were no longer present. This might have been due to a combination of concurrent associations (1) and correlations between repeated measures of the same outcome over time (2). Social support (i.e. the social and environmental circumstances influencing willingness to wait) provide a more powerful explanation for the differences in the marshmallow test than self-control. The marshmallow test may be predictive because it reflects social factors. Children who grow up in supportive environments may be more willing to delay gratification because of greater trust that events will unfold as expected and because of norms around future-oriented decisions. Experiences with...