Abstract
Aims
To highlight the importance of friendships to young people with psychosis, and the need for clinical interventions to help maintain peer relationships during illness. To structure a research agenda for developing evidence-based interventions with friends.

Method
An argument is developed through a narrative review of (i) the proven efficacy of family interventions, and (by comparison) a relative absence of friend-based interventions; (ii) the particular primacy of friendships and dating for young people, and typical effects of exclusion; and (iii) reduced friendship networks and dating experiences in psychosis, in pre-, during and post-psychosis phases, also links between exclusion and psychosis.

Results
We put forward a model of how poor friendships can potentially be a causal and/or maintenance factor for psychotic symptoms. Given this model, our thesis is that interventions aiming to maintain social networks can be hugely beneficial clinically for young people with psychosis. We give a case study to show how such an intervention can work.

Conclusions
We call for ‘friends interventions’ for young people with psychosis to be developed, where professionals directly work with a young person's authentic social group to support key friendships and maintain social continuity. An agenda for future research is presented that will develop and test theoretically driven interventions.