Commentary to: The case for investigating social context in laboratory studies of smoking Dimoff & Sayette not only make a persuasive case for taking account of social context in laboratory studies of smoking: their paper also underlines the importance of interdisciplinary perspectives on smoking behaviour, and particularly those grounded in social science and social epidemiology. John Dimoff & Michael Sayette argue that laboratory-based smoking research should ‘systematically and comprehensively’ examine social context 1. They support their argument in two ways. First, they point to the dearth of laboratory studies which include contextual factors in the study design. In their review of laboratory studies of smokers, only a handful incorporated an environmental factor—the presence or absence of another smoker—in the experimental design. The majority of these studies found that smoking behaviour was affected by the presence of ‘a peer confederate’ who was also smoking. The second source of evidence comes from a wider range of psychological studies demonstrating that ‘smoking can be influenced by the social contexts in which it occurs’. Examples are given of the influence of intimate relationships and wider social networks on smoking norms, behaviour and identities. The authors weave these two sources of evidence into a persuasive case for incorporating social context. However, what is striking from the standpoint of a researcher working outside psychology is that the case was one that needed to be made. Within the broader social science and public health fields, studies have social context as their starting point and often as their primary focus. There are therefore rich seams of research to support the author's argument and inform the deeper appreciation of context that they are seeking. For example, studies have described how the social patterning of cigarette smoking changes over time: following the introduction of the new tobacco product, prevalence is often higher among more advantaged groups (higher-educated, better-off) before social gradients flatten and smoking is associated increasingly with social disadvantage. In high-income countries where cigarette smoking first took hold, the habit is increasingly confined to poorer groups. Being a smoker has become a stigmatized identity linked inextricably to other stigmatized identities (for example, to ‘welfare mother’ in the United States and ‘chav’ in the United Kingdom) 2. Taking account of social context has also turned the spotlight on how the circumstances of people's lives, from childhood and through adulthood, shape their smoking trajectories. A disadvantaged pathway through life—poor childhood circumstances, early school-leaving, early parenthood, adult disadvantage—increases the risk of smoking. Cultural contexts matter too: in the United Kingdom, for example, South Asian women are at greater risk of social disadvantage but are less likely to smoke than white women. These broader seams of research have also shed light on intimate relationships and social networks as important contextual influences. Studies have noted how quitting can disrupt valued social relationships and the identities they sustain. It can mean a loss of shared activities and intimacy with partners, family and friends 3. As these examples suggest, Dimoff & Sayette's paper not only makes the case for investigating social context in laboratory studies of smoking; in so doing, it underlines the importance of interdisciplinary approaches to understanding smoking behaviour. Pharmacological and psychological explanations are not enough; they need to be enriched by perspectives from social science and social epidemiology. None.