In the Netherlands, the Demands and Capacities Model based treatment (DCM) has been the standard treatment for pre-school stuttering children since the late 1980s. The Lidcombe Program (LP) is a behavioural treatment program for stuttering pre-schoolers that was developed in Australia in the 1990s. Since 2000, this treatment program is also practised by a minority of Dutch clinicians. In 2007, a randomised trial in which both treatments were compared on effectiveness and costs was started. This study was named RESTART: the Rotterdam Evaluation study of Stuttering Therapy in preschool children- a Randomised Trial.In total, 199 stuttering children aged 3-6 years old and their parents participated in this study. They presented for treatment to one of the 24 participating clinicians who were qualified for both treatment methods (DCM and LP) and whose clinics were scattered over the Netherlands. To be admitted to the study, the children had to stutter at least mildly according the parent and the clinician, and children and parents had to be proficient in Dutch. Children stuttering less than six months before recruitment were excluded. Assessments were scheduled pre-treatment, 3, 6, 12 and 18 months post treatment, including audio speech samples outside the clinic, video recording inside the clinic, and questionnaires.Randomisation over treatment was stratified by age, gender, time since onset, severity of stuttering, family history of recovery/persistence from stuttering, and clinician. Every three months, meetings between the research group and all participating clinicians were held, enabling deliberation.At the conference, the results of the RESTART study regarding recovery of stuttering and cost effectiveness will be presented.