Purpose of the study Treatment resistance is a frequent situation in Obsessive Compulsive Disorder (OCD), occurring in 40–60% of patients. However, in the current literature there are only a few evidence-based options for treatment resistant patients. Pulse-loading treatment consists in a rapid titration of the pharmacological agent in the first days of treatment. A few studies suggested that this kind of titration with intravenous clomipramine could result in a greater and faster response than with a standard titration in OCD resistant patients Koran, 2006 , Grassi, 2013 . The aim of this open-label trial was to investigate the effectiveness and tolerability of a citalopram pulse-loading protocol in severe treatment-resistant OCD patients. Methods We enrolled 7 severe treatment-resistant OCD patients. Treatment-resistance was defined as non-response to two Serotonin Re-uptake Inhibitors (SRIs) trials, one antipsychotic augmentation trial and a non-response to 16 sessions of Cognitive-Behavioral Therapy (CBT) (Pallanti et al., 2002) . After a 2-week washout period from previous SRIs medications, patients were treated with intravenous citalopram starting with 40 mg for 3 days and increasing the dose up to 80 mg from the fourth day. The patients continued the treatment with 80 mg of intravenous citalopram for 18 days (a total of 21 days of intravenous treatment), then they switched to oral treatment (80 mg of oral citalopram). The patients treated with an antipsychotic augmentation, continued their antipsychotic treatment without any change during all the treatment protocol. No cognitive or behavioral psychotherapy was allowed during the intravenous or oral citalopram treatment period. We assessed treatment response using the Yale-Brown Obsessive-Compulsive Scale (YBOCS) and the Clinical Global Impression – Improvement scale (CGI-I) at baseline, week 3 and week 12. The pulse-loading protocol incorporated standard criteria for treatment response (>35% improvement in baseline Y-BOCS scores and a CGI-I of 1 or 2), partial response (≥ 25% improvement in baseline Y-BOCS scores), and non-response ( [3] . Tolerability was assessed with a clinical report form. We assessed putative cardiac side effects with an electrocardiography at baseline, after the fourth day of infusion and before switching to oral therapy, monitoring any QTc change. We also monitored sodium levels during all the course of treatment. Results During the pulse-loading treatment no patients showed significant adverse events. No patients showed clinical significant change of the QTc interval and/or of the sodium levels. Four out of seven patients had a partial or full response at the end-point (3 patients had a full response and 1 patient had a partial response) and one of these had remission. Pulse-loading treatment seemed to induce a faster improvement respect to a standard titration, since the responder patients showed a significant improvement already after 3 weeks. Three out of seven patients did not respond. Conclusion Taking into account the very limited sample size, this case series suggests that this treatment approach deserves further controlled studies.