Securitization of infectious diseases may involve suspension of ordinary human rights and liberties. In the event of an epidemic, therefore, it is important to limit the occasions upon which draconian disease control measures are implemented in the name of security. The term 'security', moreover, should not be used too loosely if it is to retain force and meaning in political discourse. It may be argued that the bar for disease securitization should be set high so that it is limited to contexts involving rapidly spreading pathogens. Such an approach, however, would rule out securitization of more slowly spreading, endemic diseases such as HIV/AIDS. An advantage of characterizing HIV/AIDS as a security threat in developing countries, where the burden of the disease is concentrated, is that this is likely to mobilize resources needed to improve the situation there. That is, if HIV/AIDS is convincingly framed as a security threat, then governments may recognize self-interested reasons to ramp up control measures. Following consideration of arguments for narrow (excluding HIV/AIDS) versus broad (including HIV/AIDS) conceptions of security, we conclude that the legitimacy of 'securitizing' HIV/AIDS ultimately turns on empirical and semantic issues, and we emphasize the importance of distinguishing (1) the nature of the threat posed by HIV/AIDS and (2) the measures required to address that threat.