Patient and public involvement (PPI) is framed as beneficial for individuals and for the health system. However, little is known about the extent of involvement, or of its impact. Based on data from Sweden, we show that apart from voting in regional elections (76%), more people reported involvement as individual patients (23%) than part of collective activities (5%) or activities relating to a citizen perspective (4%). There was no correlation between how many people participated and the estimated impact – which was generally low. More extensive involvement is thus not linked to the potential to influence decisions. We argue that to achieve the benefits associated with PPI it is crucial to understand more about people's motivation for being involved and what underlies low estimates of impact. This requires a more systematic approach to involvement, how it is evaluated and its results communicated to participants and the society. We also argue that a future challenge for the Swedish health system, and for other similar health systems, is to support long-term collective involvement in the midst of growing individualization of health services and involvement opportunities primarily intended for patients.