Health inequalities are understood to be unfair and systematic differences in health among and between social groups – differences which need to be addressed through action. These result from social and political circumstances and are therefore potentially avoidable. To address these inequalities, the relationships between the determinants of health and the health of the population have been brought to the fore so as to direct political action, which can include programmatic intervention at several levels. Despite repeated calls for more action at the structural level and despite political recognition of the importance of this type of action for reducing health inequalities (Popay, Whitehead, & Hunter, 2010), in reality, for various ideological, historical or practical reasons (Baum, 2011; Baum & Fisher, 2014), policies have more generally aimed at promoting healthy lifestyles and behaviour (e.g., the tax credit promoting physical activity for children in families). This tendency to recognize the need to act on the more structural determinants of health inequalities but to instead develop interventions targeting the more behavioural determinants of health is sometimes called ‘lifestyle drift.’ This has heightened the individualization of responsibility for health (Baum & Fisher, 2014; Baum, 2011) and in some cases, limited the reduction of inequalities or even, led to their intensification (Scott-Samuel & Smith, 2015). There is also a preponderance of policies targeting individuals and communities that are already disadvantaged rather than an attempt to reduce inequalities across the gradient. Such policies limit action that effectively reduces health inequalities throughout the population (Popay et al., 2010). Our goal is to clarify how the different broad approaches to addressing inequalities are grounded theoretically and how they affect inequalities differently. To better understand the different potential impacts of these approaches, which we briefly define in the text, we shed some light on three interrelated dimensions that are often overlooked or misunderstood.