The treatment protocol and baseline characteristics of 480 subjects with a history of repeated parasuicide recruited in five centres to a randomised therapeutic trial of manual assisted cognitive-behaviour therapy (MACT) and treatment as usual (TAU) are described. Most patients had significant anxiety and depressive disturbance with 42% having a personality disorder. Variation in service policies influenced recruitment, with earlier assessment centres seeing people with more frequent episodes of self-harm and greater parasuicide risk than later ones. Parasuicide risk was also significantly greater in those with their first parasuicide episode at an earlier age and in those with a more recent latest episode.