Introduction: Suicide is a leading cause of death and people experiencing psychosis are at increased risk of dying by suicide. The development of new therapies to reduce suicidal thoughts, plans and acts is a priority for mental health services. The Cognitive AppRoaches to coMbatting Suicidality (CARMS) project investigated the effectiveness of cognitive behavioural suicide-focused therapy for people experiencing suicidality and psychosis (CBSPp). This paper will present some preliminary data on engagement in therapy, acceptability, and effectiveness. Method: CARMS was a two-armed randomised controlled trial comparing suicide-focused therapy (CBSPp) plus Treatment As Usual (TAU) with TAU alone. Therapy was delivered over 6 months in up to 24 sessions. Participants were recruited from NHS mental health services, were 18 years old or more, were experiencing a non-affective psychosis and were experiencing suicidal thoughts, plans, and/or attempts in the three months prior to recruitment to the trial. We assessed multiple mental health outcomes at baseline, 6 and 12 months, recorded therapy attendance and the level of formulation achieved during therapy. Results: 292 people took part in the trial; 149 received Suicide-focused Therapy (CBSPp) plus TAU. The impact of the intervention on suicidal thoughts will be described. With regard to therapy attendance, a median of 16 out of 24 sessions offered were attended although the range was large. A small number of people attended no sessions. There was some evidence of a ‘dose’ response in that the more sessions received, the greater were there reductions in suicidal ideation severity. The level of formulation achieved during sessions was variable but a large majority of participants were able to discuss and formulate their suicidality in sessions. Discussion: Our suicide focused therapy was acceptable and there was a suggestion it was effective for those who engaged in therapy. Although talking about suicide was difficult and sometimes distressing, therapy attendance suggested it was a welcomed approach. Taken together with findings from qualitative work, there is value in further research to examine how CBSP can be optimised to meet service user needs and become available to those people who want it.