Obsessive-compulsive disorder (OCD) continues to present a particular challenge to clinicians. As opposed to other anxiety disorders and, to some degree, to depression, the results of therapies, be they pharmacologie or psychological, are at best less than optimal. Indeed, when one reads articles reporting randomized controlled trials, patients are said to be responders when a 35% reduction of symptoms occurs (as if reducing rituals from 6 to 4 hours were clinically meaningful). Moreover, when one takes into account those who drop out of studies because of medication side effects or because of fear in the exposure-response prevention (ERP) studies-often in the 25% to 30% range-and add to those numbers the nonresponders, then we are looking at a 35% to 50% response in about 50% of patients. Additionally, few patients attain full remission-hardly satisfactory outcomes! The one redeeming finding is that most gains achieved by cognitive-behavioural therapy (CBT) seem to be stable. The longest study to date spans 7 years but has few subjects (1). In it, the 41% improvement was maintained (45% at follow-up). Hence the continued effort to develop new strategies. Foa reports an intensive regimen of daily, prolonged exposure sessions (over 2 hours) for 3 weeks (2). The dropout rate was extremely high in the New York site (over 40%), as opposed to the Philadelphia site, which may indicate that in Foa's clinic the cognitive preparation for treatment reduced apprehension and induced better compliance. Dr Foa's comparator to ERP was clomipramine, which we know has a high side effect profile that may explain patient resistance. The ERP therapy was superior to clomipramine alone, and the combination of behaviour therapy and medication did not, in that study, improve outcome. Foa feels that ERP was so powerful that there was little room for the medication to show added improvement. However, in children and adolescents, the combination of CBT and sertraline has been shown to be superior to either treatment alone (3). Another strategy has been to add the components sequentially. For example, Kampman added 12 sessions of CBT to continued treatment with fluoxetine in nonresponders (defined as those showing less than 25% improvement after 12 weeks of medication) (4). There was a 41% improvement rate in these resistant patients. A recent review and 2 case illustrations have outlined in which patients the combinatory approaches may be best indicated (5). In his review in this issue, Abramowitz summarizes the comparative results of CBT and ERP and shows that the indispensable component of the psychological approaches seems to be exposure, even in the soi-disant pure cognitive group (individuals often expose in imagination and indeed, at times, spontaneously in vivo without therapist intervention). Abramowitz also mentions that cognitive therapists have developed strategies that look more deeply into the schemas of patients-a core belief of enhanced responsibility for the patient's own and others' well-being (6), an increased sense of vulnerability (7), and thought-action fusion (IfI think it, it means I want to do it) (8). Integrating ERP, traditional CBT, and schema-focused cognitive therapy has been proposed as a strategy to enhance response in treatment-resistant patients (9). Subtypes identified according to symptoms (such as checking, washing, or symmetry) may also explain some of the differential responses encountered in treatment (10) Augmenting regimens have also been proposed in pharmacology: in Blier's review, he states that the most promising approach seems to employ the atypical antipsychotics. …