In October, 2010, a 76-year-old woman born in the USA presented to our emergency department complaining of a severe sore throat and painful swallowing. She reported having had a non-productive cough without haemoptysis 1 month previously. The cough resolved, but a sore throat then developed. At the time she was assessed by her general practitioner and specialists in the emergency and otolaryngology departments, who diagnosed pharyngitis and recommended conservative management. Her odynophagia progressed until she was unable to eat solid foods, prompting her to present to our hospital. She denied fevers, chills, night sweats, myalgias, dysphonia, dyspnoea, or dysphagia but did report a 7 kg weight loss. She also reported occasional tobacco and alcohol use. On examination she was afebrile and thin but appeared well. She had dry mucous membranes and mild erythema of the posterior pharynx without tonsilar enlargement or exudate. No oral thrush, cervical lymphadenopathy, rales, splenomegaly, or rash were noted. Leucocyte concentration was 7·9 × 109/L with a predominance of neutrophils. She was admitted for further assessment. Rapid streptococcal antigen test, monospot test, human immunodefi ciency virus (HIV) antibody test, and rapid antigen test for infl uenza A and B were negative. A temperature of 38·7°C prompted a chest radiograph, which showed bilateral airspace disease and an area of dense consolidation in the right upper lung with a rounded lucency likely to be cavitation (fi gure A), subsequently confi rmed on chest CT (fi gure B). CT of her neck showed non-specifi c epiglottis oedema. Purifi ed protein derivative skin test was negative. A few acid fast bacilli were present in the sputum, and she was started on empiric therapy for presumed pulmonary tuberculosis. DNA probe performed on sputum confi rmed Mycobacterium tuberculosis which was sensitive to all antituberculous drugs tested. Subsequently, her fever and odynophagia resolved. Laryngeal tuberculosis was the most likely cause of our patient’s sore throat as evidenced by her clinical symptoms and presumed active pulmonary tuberculosis, the non-specifi c oedema of the epiglottis noted on CT, and the resolution of her sore throat and odynophagia with antituberculous therapy. However, a defi nitive diagnosis cannot be made without laryngoscopy and histopathological confi rmation; laryngoscopy was deferred because of the substantial infectious risk associated with the procedure in the setting of active tuberculosis. She was discharged with instructions for tuberculosis testing of close contacts and plans for direct observed therapy. At last contact in February, 2011, she was well, with no recurrence of symptoms. Laryngeal tuberculosis, which accounted for 25% of all cases of tuberculosis in the early 20th century, now accounts for less than 1% of cases and is typically associated with pulmonary infection. Whereas extrapulmonary tuberculosis occurs frequently in persons with HIV, cases of laryngeal tuberculosis have been reported in immunocompetent persons. Our patient presented with odynophagia, a common presenting complaint in patients with laryngeal tuberculosis, but she denied hoarseness, the most common presenting symptom. She denied any history of tuberculosis or known contact with persons with tuberculosis, but she did report having worked as a home health aide during the 1970s. Laryngeal tuberculosis results from either direct bronchogenic spread from infected sputum or, less commonly, haematogenous spread. Laryngeal tuberculosis is highly infectious, and a delay in diagnosis can lead to disease progression and longer exposure to contacts. Therefore, physicians should consider laryngeal tuberculosis in the diff erential diagnosis of patients presenting with a persistent sore throat and odynophagia, especially in patients such as ours who also report weight loss. A chest radiograph should be considered before laryngoscopy to minimise unnecessary risk to healthcare providers.