Abstract Aims To determine the clinical outcomes of an intensity-modulated radiotherapy technique for total mucosal irradiation (TM-IMRT) in patients with head and neck carcinoma of unknown primary (HNCUP). Materials and methods A single-centre prospective phase II trial design was used in two sequential studies to evaluate TM-IMRT for HNCUP. Patients were investigated for primary tumour site using examination under anaesthetic and biopsies, computed tomography ± magnetic resonance imaging (MRI) or 18-fluorodeoxyglucose positron emission tomography-computed tomography (PET-CT). Patients received IMRT to the potential primary tumour sites and elective cervical nodes. Concomitant chemotherapy was used in patients who received primary radiotherapy or those with nodal extracapsular extension. Results Thirty-six patients with HNCUP were recruited; 72% male. Twenty-five patients (69.4%) had p16-positive disease. Two year mucosal and local nodal control rates were 97.1% (95% confidence interval 91.4–100) and 89.8% (78.4–100), respectively. One mucosal primary was detected 7.3 months after TM-IMRT and three patients died from recurrent/metastatic squamous cell carcinoma of the head and neck. Twelve patients (33%) developed grade 3 (Late Effects in Normal Tissue-Subjective, Objective, Management and Analytical; LENT-SOMA) dysphagia with a 1 year enteric tube feeding rate of 2.7%. The high-grade subjective xerostomia rate (LENT-SOMA) at 24 months after IMRT was 15%. Conclusions At a median follow-up of 36.1 months, the use of TM-IMRT was associated with good local control. Toxicity was comparable with previously reported TM-IMRT regimens encompassing similar mucosal volumes.