This article describes the sequence of acute clinical deterioration seen in a head and neck oncology patient who developed serotonin syndrome perioperatively. It highlights the clinical dilemma that can be encountered when a septic picture masks the onset of serotonin syndrome and reinforces the importance of awareness of the potential interactions and side effects associated with drugs that surgeons prescribe. We discuss the pathophysiology, causal factors, clinical presentation and diagnosis of serotonin syndrome as well as highlighting some of the dilemmas that this condition presents in the surgical setting. Ann R Coll Surg Engl Case historyMr K is a 77-year-old patient admitted for a hemimaxillectomy for a primary squamous carcinoma of the palate. His co-morbidities were insulin-dependant diabetes, hypertension, mild depression and chronic lymphoedema. Preoperatively, all blood tests were normal. His diabetic control was reviewed and optimised. A week after the uneventful operation, the patient developed acute depression for which fluoxetine was prescribed. Three days after starting the selective serotonin reuptake inhibitor (SSRI), Mr K developed a neutrophilia and swinging pyrexia.A swab of a gangrenous toe grew methicillin-resistant Staphylococcus aureus (MRSA) but three blood cultures were negative. As there was radiological and clinical evidence of early osteomyelitis, intravenous vancomycin and rifampicin were started. When no improvement was seen after four days and a blood culture grew MRSA, microbiologists advised to start the anti-MRSA antibiotic linezolid. Over the next 48 hours Mr K became increasingly agitated and confused. His pyrexia persisted and was accompanied simultaneously by hypertension, tachycardia, rigors, ataxia, agitation and insomnia, which all worsened over the next 72 hours. A neurological examination revealed generalised hypertonicity and weakness of all limbs with intact sensation in the upper limbs and an unchanged diabetic neuropathy in his lower limbs. Coordination and proprioception could not be tested due to the patient's confusion but there were no signs of cerebrovascular events. Inducible clonus was present in both feet and reflexes were generally brisk.