This article raises ethical issues that result if individuals, during the course of research to measure the prevalence of drug-resistant tuberculosis, are identified with the disease but are not provided with or referred for appropriate treatment. It draws attention to and applauds recently published World Health Organization guidelines on the subject. Questions posed are: 1) Should treatment be provided for individuals identified through surveillance projects for MDR-TB, whatever their purpose (specific research or 'routine' national prevalence studies)? 2) If treatment availability is a problem, who is responsible for assuring this?