Knowledge of HIV and AIDS is widely seen as an important determinant o
f anxiety about such infection in health care workers. However, existi
ng measures of this knowledge suffer from a number of methodological p
roblems and few demonstrate adequate reliability and validity. This pa
per documents the development of a new measure detailing its reliabili
ty and validity. Knowledge, and also confidence in knowledge, were ass
essed across seven domains of relevant information (e.g. epidemiology,
personal risk, symptomatology) in one non-expert group (non-medical u
ndergraduate psychologists) rand in three groups with greater expertis
e (nursing students and third year and final year medical students). T
he measure was shown to be reliable and to discriminate between the fo
ur groups in a manner consistent with the level of education received
by each group and in terms of bath the scab scores for the seven domai
ns and the individual items. In general, final year medical students w
ere shown to be more knowledgeable, and more confident in their Knowle
dge, than the other groups. The non-medical undergraduates showed the
lowest level of knowledge. In general, the groups appeared under-confi
dent in their knowledge about HIV/AIDS. This was interpreted in terms
a members of these groups exhibiting caution and the perceived fluctua
ting nature of the HIV knowledge base. Implications for the training o
f health care workers in relation to HIV/AIDS are discussed.