This paper takes a new look at the long-standing debate about whether the capability approach is too indi­vidualistic and how best to take account of collectivity. The debate has been summarised several times (Robeyns, 2005; Alkire, 2008) but it is still ongoing with some recent contributions to the empirical as well as the conceptual level (Godfrey-Wood & Mamani-Vargas, 2016; John B. Davis, 2015; Hall, 2016). This paper broadens the view by going beyond the papers that stirred the question of individualism and tracing more broadly various strands of the relevant literature. In order to get a grip on the large range of papers a classification scheme is proposed first. The literature can be classified according to how it conceives collectives. Some understand collective in a descriptive way, assigning membership to all people who share certain characteristics regardless of their own view. For example, Stewart’s (Stewart, 2005; 2010) research on ‘group capability’ and ‘horizontal inequality’ shows impressively that group deprivations are lasting longer than individual deprivations by relying on statistical analysis and external descriptions of deprived groups. In contrast to that the well-known notion of ‘collective capabilities’ (Evans, 2002; Ibrahim, 2006) refers to groups based on voluntary and conscious membership. In their case, people are committed to the group in question and usually know its members personally. Thus, this kind of collective is probably much smaller than that referred to by Stewart. A second criterion for classifying the strands of literature derives from the goals the groups are pursuing: The capability approach is in general mainly concerned with well-being. The literature on collectives point to the positive instrumental role groups play in promoting the well-being of disadvantaged people. Sen (1985c) holds that the capability approach is especially apt for assessing well-being, but that humans usually have goals beyond their own well-being. These ‘agency goals’ differ widely from person to person as they encompass commitments to groups, political convictions and ‘the morals’ more generally. According to Sen (1999d) commitments may drive a wedge between a person’s choice and her well-being, thus it is important to distinguish well-being and agency. Collectives play an important role in exercising agency hence contrib­uting to the constitutive role of democracy by providing spaces for public deliberation. Some collectives mainly aim at improving the well-being of their members, others mostly at strengthening their agency. This is the second classification criterion. The paper then surveys the literature and distinguishes five main strands classified according to the cri­teria. Collective capabilities are but one way to conceive groups from a capability perspective. The classification frame can also be applied to the participatory research process followed in the project RE-InVEST. Some problems of doing participatory research on the subject of social policy and within a huge third-party-funded project are highlighted. The conclusion summarises the main findings and shows how the various approaches complement each other in explaining various aspects of reality.