Abstract Background Many persons with autism spectrum disorder (ASD) are treated in long-term specialised care. In this population, suicidal behaviour troubles patients, families, and specialists in the field because it is difficult to treat. At present, there is no documented effective therapy for suicidal behaviour in ASD (Autism Research 7:507-521, 2014; Crisis 35:301-309, 2014). Dialectical Behaviour Therapy (DBT) is an efficacious treatment programme for chronically suicidal and/or self-harm behaviour in patients with Borderline Personality Disorder (J Psychiatry 166:1365-1374, 2014; Linehan MM. Cognitive behavioural therapy of borderline personality disorder. 1993). This study will evaluate the efficacy of DBT in persons with ASD and suicidal/ self- destructive behaviour in a multicentre randomised controlled clinical trial. Method One hundred twenty-eight persons with autism and suicidal and/or self-harming behaviour will be recruited from specialised mental healthcare services and randomised into two conditions: 1) the DBT condition in which the participants have weekly individual cognitive behavioural therapy sessions and a 2.5 h skills training group session twice per week during 6 months, and 2) the treatment as usual condition which consists of weekly individual therapy sessions of 30–45 min with a psychotherapist or social worker. Assessments will take place at baseline, at post-treatment (6 months), and after a follow-up period of 12 months. The mediators will also be assessed at 3 months. The primary outcome is the level of suicidal ideation and behaviour. The secondary outcomes are anxiety and social performance, depression, core symptoms of ASD, quality of life, and cost-utility. Emotion regulation and therapeutic alliance are hypothesised to mediate the effects on the primary outcome. Discussion The results from this study will provide an evaluation of the efficacy of DBT treatment in persons with ASD on suicidal and self-harming behaviour. The study is conducted in routine mental health services which enhances the generalisability of the study results to clinical practice. Trial registration ISRCTN96632579. Registered 1 May 2019. Retrospectively registered.