Abstract Anorexia nervosa is difficult to treat and no treatment is supported by robust evidence. As it is uncommon, it has been recommended that new treatments should undergo extensive preliminary testing before being evaluated in randomized controlled trials. The aim of the present study was to establish the immediate and longer-term outcome following “enhanced” cognitive behaviour therapy (CBT-E). Ninety-nine adult patients with marked anorexia nervosa (body mass index ≤ 17.5) were recruited from consecutive referrals to clinics in the UK and Italy. Each was offered 40 sessions of CBT-E over 40 weeks with no concurrent treatment. Sixty-four percent of the patients were able to complete this outpatient treatment and in these patients there was a substantial increase in weight (7.47 kg, SD 4.93) and BMI (2.77, SD 1.81). Eating disorder features also improved markedly. Over the 60-week follow-up period there was little deterioration despite minimal additional treatment. These findings provide strong preliminary support for this use of CBT-E and justify its further evaluation in randomized controlled trials. As CBT-E has already been established as a treatment for bulimia nervosa and eating disorder not otherwise specified, the findings also confirm that CBT-E is transdiagnostic in its scope.